300.1. Scope. Part (A) of this section indicates that the rules of Art. 300 apply to all installations of the wiring methods covered by this article and the remainder of Chap. 3. As is generally the case, clearly not all the general requirements in Art. 300 apply to specialized installations, such as remote-control circuits, to signal circuits, to low-energy circuits, to fire-protective signaling circuits, and to communications systems. The wording “unless modified by other articles” is intended to convey that idea.
Part (B) establishes what amounts to the boundary line between the NEC and UL or other equipment testing labs, at least with respect to this article, which has no “construction” part. In fact the NEC contains requirements throughout that are effectively aimed at the equipment manufacturers, in the process of setting policy around how products need to perform in order to be used safely, based on a consensus vote of a code-making panel (CMP). For only one of countless examples, the reason there is foam-core (“nonhomogeneous”) PVC conduit available is that a code-making panel voted it in [352.10(G)], but then also limited its use to underground applications. The reason that UL, with well-earned pride, points to its representation on all the code-making panels is that it is then in a position to make sure that its product standards correctly implement the policy decisions reached through the NFPA consensus process. This rule clearly states that there is no intent, or permission, to apply Code rules within Art. 300 to the interior wiring of equipment to be connected. That is the province of the testing labs. If a product is “listed,” it has been tested and found to be essentially free from hazards. Generally, the equipment is tested for use in accordance with the Code, which includes any Code-prescribed construction or performance requirements.
In part (C) of this section, a table is provided to allow proper selection of standard trade size raceways and tubing whether applying the Code in the United States or abroad. Simply determine the needed raceway size, in either system, and use Table 300.1(C) to establish the correct metric or English system size equivalent. The table also establishes that the dimensions are not real dimensions, but only to be used for identification purposes. Generally throughout this book we are following the NEC style, and therefore if we are referring to what is known on the street as “3-in. conduit” we refer to it here, in the customary NEC practice of listing the metric size first and the English unit following in parentheses, as “metric designator 78 (trade size 3) conduit.”
300.3. Conductors. Part (A) requires that single conductors described in Table 310.13 (A) must be used only as part of one of the wiring methods covered in Chap. 3. This basically means that the various insulated conductors recognized in Table 310.13(A) may not simply be strung overhead, without benefit of being incorporated into a cable assembly or otherwise protected and supported by conduit or tubing. The Exception to part (A) recognizes the permission given in 225.6, where individual conductors are permitted to be run as “open conductors” in overhead feeders and branch circuits installed outdoors, as well as in festoon lighting.
Part (B) requires that all conductors of the same circuit—including the neutral and all equipment grounding conductors—must be run in the same raceway, cable tray, trench, cable, or cord. Part (B)(1) recognizes the use of separate raceways and cables, where circuits are made up of multiple (two or more) sets of conductors or cables in parallel. The exception that follows correlates with 300.5(I) Exception No. 2 for isolated phase installations. Part (B)(2) in that rule notes those very specific and unusual sections of the NEC, where an equipment grounding conductor may be run separately from the other conductors of the circuit such as for DC circuits and for retrofits under the provisions of 250.130(C).
Part (B)(3) requires taking steps to prevent induced currents where nonmetallic or nonmagnetic sheathed conductors are run through metallic enclosure walls with magnetic properties as covered in 300.20(B). The second sentence gives a similar warning for Type MI cables and references 332.31, which addresses the concern for induced currents. Both of those other rules discuss methods for preventing induced currents.
Part (4) of 300.3(B) permits limited use of a pull box equipped with a terminal block for the connection of the system neutral, as the point of origin for branch circuit neutral conductors. That is, a properly sized neutral is run to the pull box—which is connected by an auxiliary gutter to a column-width panelboard—from the panelboard, and the individual branch-circuit neutrals may be run from the pull box and need not go back to the panelboard where the hot conductors originate. This saves space within the panelboard, but is only permitted for column-width panelboards connected by an auxiliary gutter to a pull box that is manufacturer-equipped with a neutral terminal block. Note that there is no inductive heating in this arrangement and it is arguably permitted even without this permission. Every ampere that comes up the auxiliary gutter from the panel on a branch-circuit conductor is equaled by an ampere of current moving down the same gutter space to the panel over one of the feeder conductors.
Part (C) covers system separations under ordinary conditions. The first topic is conductors operating on different voltages, but not over 600 V. This rule clearly allows conductors of different voltage systems, and whether dc or ac, to occupy the same raceway, cable assembly, or enclosure. The only limitation here is that all the conductors must be insulated for the maximum voltage that will be present in the common location. For example, nonpower-limited 6-V circuit conductors running to a valve motor can run in the same raceway as 277-V circuit wiring to a duct heater, if both sets of conductors have 600-V (technically 277-V) insulation.
This brings up Fig. 300-1. Whoever wired this furnace must have read this section, because when he wanted to get his 24-V thermostat circuit neatly down to the burner control, he carefully ran 600-V THHN down the EMT riser along with the 120-V circuit for the burner motor, making the transition to thermostat wire where the THHN poked out of the upper end of the riser. A nice neat job, and a serious code violation. The thermostat circuit is a Class 2 power-limited circuit, covered under the enhanced system separation rule found in 725.136(A). This comes up so often that the NEC now has a note at this location. A Class 2 circuit is a special condition, covered in Chap. 7 of the NEC. As such, the rules in Chap. 7 automatically supersede or modify information in Chaps. 1 through 4 of the NEC, as per 90.3. And for Class 2 wiring, it is forbidden to ever rely on insulation alone to define a circuit separation. These systems are presumed to be incapable of fire and electrocution hazard by virtue of the limitations built into their power supplies and the only way to be really sure of this is to keep them out of common raceways and enclosures, except under extremely well-controlled situations. There is even a special exception to the rule forbidding cables to be attached to electrical raceways [300.11(B)(2)] that allows the Class 2 wiring to be attached to the outside of the raceway, and that would have been the way to wire the oil burner in Fig. 300-1.
Fig. 300-1. The 14 AWG THHN leaving the box from the EMT riser and connected to 18-2 thermostat wire for a 24-V Class 2 circuit does comply with 300.3(C)(1), but violates 725.136(A).
This raises the question as to why the exception for a Chap. 6 rule is in this location, when there is only a note for the Chap. 7 rule. This exception is for solar photovoltaic applications covered in 690.4(B). Photovoltaic source and output circuits are DC circuits, often without overcurrent protection because the only possible current is the output of the module. Ever since Art. 690 entered the NEC in the 1984 edition, these circuits have been segregated from normal power applications; and there is no problem with technical merit. In fact, however, this exception should be removed and converted to a note as well. Back in the 1984 NEC it was commonly believed that these exceptions in Chaps. 1 to 4 were necessary, and this quite simply has not, as of yet, been attended to.
Part (C)(2) of 300.3 states that conductors operating at more than 600 V must not occupy the same equipment wiring enclosure, cable, or raceway with conductors of 600 V or less. There are five conditions where this can be relaxed; however, note that none of these involve a common raceway. Race-ways remain fully segregated, but there are limited applications where having both systems in a common enclosure is essential. Part (C)(2)c is intended to apply to enclosures, not raceways, such as used for high-voltage motor starters, permitting the high-voltage conductors operating at over 600 V to occupy the same controller housing as the control conductors operating at less than 600 V (Fig. 300-2).
Fig. 300-2. Control wires for high-voltage starters may be used in the starter enclosure, but not in raceway with power conductors. (Sec. 300.3.)
The last sentence in part (C)(2) is intended to prohibit unshielded conductors (now limited by 310.6 to 2.4 kV) from occupying the same enclosure, raceway, or cable, unless the actual voltage carried is the same. Now that the allowable voltage for unshielded conductors has been reduced from 8 kV, the application of this in new construction is limited, but there are a lot of higher voltage unshielded conductors still in use. Some feel that the normal voltage discharge and leakage current will be increased where a difference of potential exists between two unshielded conductors. It is theorized that the result of the increased discharge and leakage will be premature insulation failure. Although the assertion of premature failure has not been determined empirically, there is no data to suggest otherwise. Any problem can be avoided by simply complying with this requirement and using separate raceway systems.
300.4. Protection Against Physical Damage. This rule presents a general requirement for ensuring that conductors are properly protected were “subject to physical damage.” Common sense must be applied and the normal operation of the facility in question must be considered. If it appears that, during normal operations, a given conductor or cable is likely to be damaged, then physical protection in the form of a raceway sleeve, an enclosure, kickplate, etc., must be provided. This requirement is a catchall. That is, the specific situations spelled out in the subsequent parts of 300.4 must be provided with the physical protection prescribed, under the circumstances described. But, in any other location where conductors or cables are exposed to damage, suitable remedies must be employed to satisfy this most basic requirement.
Part (A) gives the rules on protection required for cables and raceways run through wood framing members, as shown in Fig. 300-3. Where the edge of a hole in a wood member is less than 1¼ in. (32 mm) from the nearest edge of the member, a
-in. (1.6-mm)-thick steel plate must be used to protect any cable or flexible conduit against driven nails or screws. The same protection is required for any cable or flexible conduit laid in a notch in the wood. But, as given in Exception No. 1, rigid metal conduit, electrical metallic tubing (EMT), intermediate metal conduit (IMC), and PVC conduits do not require such protection. And, Exception No. 2 allows for the use of steel plates that are less than-in. thick where the plate is “listed and marked” as providing “equal or better protection” from nails, etc.
Clearance must be provided from the edge of a hole in a wood member to the edge of the wood member. The NE Code requires only 1¼ in. (32 mm). This permits realistic compliance when drilling holes in studs that are 3½ in. (89 mm) deep. It also was taken into consideration that the nails commonly used to attach wall surfaces to studs are of such length that the 1¼-in. (32-mm) clearance to the edge of the cable hole affords entirely adequate protection against possible penetration of the cable by the nail.
Figure 300-4 shows typical application of cable through drilled studs, with holes at centers and adequate clearance to the edge of the stud. Figure 300-5 shows an objectionable example of a drilled hole, violating the rule of part (2) of this section, which warns against “weakening the building structure.” Figure 300-6 shows an acceptable way of protecting cables run through holes in wood members.
Fig. 300-3. Holes in wood framing must not weaken structure or expose cable to nail puncture. (Sec. 300.4.)
Fig. 300-4. Holes or notches in joists and studs must not weaken the structure of a building. (Sec. 300.4.)
Fig. 300-5. Excessive drilling of structural wood members can result in dangerous notching (arrow) that weakens the structure, violating Sec. 300.4(A)(2). (Sec. 300.4.)
In part (B), the rules on installations through metal framing members apply to nonmetallic-sheathed cable and to electrical nonmetallic tubing (ENT). Part (1) of 300.4(B) applies to NM cable run through slots or holes in metal framing members and requires that such holes must always be provided with bushings or grommets installed in the openings before the cable is pulled. But that requirement on protection by bushings or grommets in the holes does not apply to ENT where run through holes in metal framing members.
Part (2) applies to both NM cable and ENT and requires that the cable or tubing be protected by a steel sleeve, a steel plate, or a clip when run through metal framing members in any case where nails or screws might be driven into the cable or tubing.
Part (C) requires cables and raceways above lift-out ceiling panels to be supported as they are required to be when installed in the open. They may not be treated as if they were being run through closed-in building spaces or fished through hollow spaces of masonry block.
Part (D) requires cables and raceways run along (parallel with) framing members (studs, joists, rafters) to have at least a 1¼-in. (32-mm) clearance from the nearest edge of the member; otherwise the cable or raceway must be protected against nail or screw penetrations by a steel plate or sleeve at least-in. (1.6 mm) thick. Part (D) exists because of many persistent reports of nail and screw penetrations of both metallic and nonmetallic cables and raceway. The rule applies to both exposed and concealed locations. Exception No. 1 excludes IMC, rigid metal conduit, rigid nonmetallic conduit, and EMT from the rule. The rule will apply to Romex (Type NM), BX (Type AC), flexible metal conduit, ENT, Type MC cable, and all other cables and raceways, except those excluded by Exception No. 1. (Fig. 300-3, bottom). Exception No. 2 excludes from this rule concealed work in finished buildings and finished panels in prefab buildings, where cables may be fished. Exception No. 3 recognizes the use of a “listed and marked” steel plate that is less than-in. thick as providing the required protection.
Fig. 300-6. Steel plates are attached to wood structure member to protect cable from penetration by nail or screw driven into finished wall, where the edge of the cable hole is less than¼ in. (31.8 mm) from the edge of the wooden member. (Sec. 300.4.)
Part (E), on wiring under roof decking, is new in the 2008 NEC and it requires care in interpretation in order to achieve the intended outcome. It requires all wiring, whether cables or raceways (subject to an exception for rigid and for intermediate metal conduits) to be mounted far enough below “metal-corrugated sheet roof decking” so subsequent reroofing and repairs don’t end up impaling the wiring with the long screws commonly used to hold down the new surface, as explained in the note. This change follows actual loss experience. Review Fig. 300-7 to understand the interpretive issue.
Fig. 300-7. The new rule to avoid damage from roof repair requires careful interpretation in order to be applied to achieve the intended effect. [Sec. 300.4(E).]
The literal text calls for 38 mm (1½ in.) spacing between the “nearest” outside edge of the wiring method (meaning nearest to the roof decking) and the “nearest” edge of the roofing (meaning nearest to an observer on the floor). This puts the wiring well below the lowest part of the corrugations, which is appropriate because subsequent roofing activities proceed blindly, with screws sized to penetrate to a bottom of a corrugation as easily as an upper portion. However, the language chosen does not clearly convey the intent. If the wiring is placed in the hollow, especially on top of bar joists, it will likely be damaged by the drill-point screws used to apply the roofing. Review this carefully with the inspector.
Although the exception waives the rule for installations using heavy-wall steel (rigid or intermediate) conduit, take care in using this exception. Tests run by UL on a variety of wiring methods to determine their susceptibility to penetration under conditions covered by the related rules in 300.4(D) showed that even heavy wall conduits were no match for drill-point screws. Interestingly, only unrestrained steel-armored cables that could roll out of the way typically survived the testing, a result not reflected in the Code.
Part (F) presents protection requirements where cables are run in “shallow grooves.” This rule calls for the same-in (1.6-mm)-thick steel kick plate, sleeving, or the equivalent. Alternately, one may locate the cable with at least 1¼ in. (32 mm) of “free space” for the cable’s entire concealed length. The “free space” is to be measured from the top of the groove to the top of the cable, not to the bottom of the groove. This rule is aimed at applications with exposed beam construction where grooves or channels are cut in the beams for supply conductors to lighting fixtures or ceiling fans.
Part (G) requires insulated bushings on raceway entries to enclosures if wiring 4 AWG or larger will be installed. The insulated surface must be rounded. Nonmetallic conduit terminal adapters are insulating, but the edge of the open throat is not always as rounded as perhaps it should be, leaving room for interpretation. It is important to remember that this is a rule that applies to wiring being pulled into raceways, and therefore there is no requirement to put an insulating bushing on the end of a SE cable connector, for example. If an enclosure has a molded hub, it will be sufficiently rounded on the inside and no additional protection is required.
300.5. Underground Installations. This section is a comprehensive set of rules on installation of underground circuits for circuits up to 600 V. (Higher-voltage circuits must satisfy 300.50.) Table 300.5 in the Code book establishes minimum earth cover needed for specific conditions of use. Figure 300-8 shows cover requirements for rigid metal conduit and IMC. Figure 300-9 shows the basic depth requirements for the various wiring methods. In all instances, “cover” is not the depth of the trench. It is the depth of the trench plus the diameter of the wiring method, which equals the amount of dirt on top of the wiring method when the job is complete.
Because Table 300.5 does not specifically mention EMT, it could be taken to indicate that the NEC does not recognize EMT for underground use. But 358.10(B) does recognize EMT for direct earth burial if “protected by corrosion protection and judged suitable for the condition”, and so does UL, with this stipulation: “In general, electrical metallic tubing in contact with soil requires supplementary corrosion protection.” Note that such protection is not literally mandatory in all instances, but it is close. The UL note means to indicate that EMT might be buried without a protective coating (like asphalt paint) where local experience verifies that soil conditions do not attack and corrode the EMT. On the other hand, UL does not evaluate supplementary corrosion protection on this product with respect to the corrosive influences of soils, so the inspector is on his own in choosing to permit this. Supplementary corrosion protection must always be applied unless there is a solid local record of positive experience, which is very unusual. On balance, EMT should not be used for direct burial. Refer to the more extensive discussion at 358.10 in this book for additional information on this point.
Figure 300-10 shows modifications of basic cover requirements. If a 2-in. (50-mm)-thick or thicker concrete pad is used in the trench over an underground circuit other than rigid metal conduit or IMC, the basic burial depth in Table 300.5 may be reduced by 6 in. (150 mm). It should be noted that the intent here is for the concrete pad to be in the trench, right over the cable or raceway. The wording must be taken to mean that it may not be a walk or other concrete at grade level. And the burial depth may not be reduced by more than 6 in. (150 mm) no matter how thick the concrete pad is. This rule is at odds with Note 3 to Table 300.50, where burial depth for high-voltage circuits may be reduced “6 in. (150 mm) for each 2 in. (50 mm) of concrete” or equivalent protection in the trench over the wiring method (other than rigid metal or IMC). Remember this when installing underground runs of conductors and cables that are rated over 600 V. For those installations, compliance with 300.50, not 300.5, must be ensured.
Note that in Table 300.5, rigid metal conduit or IMC that is buried in the ground must have at least a 6-in. (150-mm)-thick cover of earth or earth plus concrete—even if it has a 2-in. (50-mm)-thick concrete pad over it. But rigid metal conduit, IMC, or other raceways may be installed directly under a 4-in. or thicker exterior slab that is not subject to vehicle traffic, without any need for earth cover. Given the fact that rigid metal conduit or IMC may be laid directly on the ground (which supports it for its entire length) and would not necessarily require any concrete cover, there is no reason why it cannot be laid on the ground or flush with the ground and covered with at least 4 in. (102 mm) of concrete. (See Fig. 300-8.)
Fig. 300-8. These are the details involved with the use of rigid metal conduit and IMC underground.
Fig. 300-9. These are the basic burial depths, but variations are recognized in Table 300.5 for certain conditions. (Sec. 300.5.)
300.5only applies to “underground installations” and is not applicable if the conduit is laid directly on the ground. No Code rule prohibits conduit laid on the ground, provided the conduit is “securely fastened in place” (300.11) and is not exposed to physical damage such as vehicular traffic, and many such installations have been made for years. But, when conduit is installed in the ground, there is serious concern about damage due to digging in the ground, which 300.5 addresses.
As shown in Fig. 300-11, Table 300.5 recognizes that raceways run under concrete slabs at least 4-in. (102 mm) thick or under buildings have sufficient protection against digging and are not required to be subject to the burial-depth requirements given in the top line of Table 300.5. Where raceways are so installed, the rule requires that the slab extend at least 6-in. (150 mm) beyond the underground raceway, as follows:
1. Any direct burial cable run under a building must be installed in a raceway, as required by 300.5(C), and the raceway may be installed in the earth, immediately under the bottom of the building, without any earth cover.
2. Any direct buried cable under a slab at least 4-in. (102 mm) thick and not subject to vehicles is subject to the 18-in. (450-mm) minimum burial-depth requirement of Table 300-5. The reason for the equivalence in cover between the 2-in. (50 mm) concrete-in-trench rule and the 4-in. (102 mm) concrete slab rule is that when the concrete is buried in the trench an excavator will recognize it as a protective structure. The surface slab may not be easily recognized as performing an additional protective function.
Fig. 300-10. Concrete pad “in trench” permits what amounts to a 6-in. (150-mm) reduction of burial depth for circuits up to 600 V for other than RMC, IMC, and applications covered in column 5. (Sec. 300.5.)
Fig. 300-11. Table 300.5 eliminates burial-depth requirements for direct buried raceways under specified conditions. (Sec. 300.5.)
Figure 300-12 shows the mandatory 24 in. (600 mm) of earth cover given in Table 300.5 for any wiring methods buried under public or private roads, alleys, driveways, parking lots, or other areas subject to car and truck traffic. A minimum earth cover of 24 in. (600 mm) is required for any underground cable or raceway wiring that is installed under vehicle traffic, regardless of concrete encasement or any other protective measure. This requires the minimum 2-ft (600-mm) earth cover for wiring under the designated areas, including driveways and parking areas of private residences. The minimum earth cover for cables and raceways under driveways and parking areas for one- and two-family dwellings is only 18 in. (450 mm).
Fig. 300-12. All wiring methods must be at least 2 ft (600 mm) under vehicular traffic. (Sec. 300.5.)
Table 300.5 (second vertical column from right) gives limited use of lesser burial depth for the residential circuits described, as shown in Fig. 300-13. Any GFCI-protected residential “branch circuit” not over 120 V and protected at 20 A or less may be buried only 12 in. (300 mm) below grade, instead of, say, 24 in. (600 mm), as required for Type UF cable for any nonresidential use or for a residential “feeder.”
Fig. 300-13. This is OK only for a residential branch circuit rated not over 20 A and protected by a GFCI circuit breaker. (Sec. 300.5.)
Figure 300-14 shows three other special conditions for burial depth. Table 300.5 (vertical column at right) recognizes reduced burial depth for low-voltage landscape lighting circuits and supply circuits to lawn sprinkler and irrigation valves, as shown in Fig. 300-15. This recognizes the reduced hazards and safety considerations for circuits operating at not more than 30 V. However, such systems must be supplied by Type UF cable or other “identified” cable or raceway. Such identification is generally required to be marked by the manufacturer. On the cables and conductor, the marking should include a “W” indicating suitability for use in a “wet location” to ensure compliance with part (B).
Part (B) of 300.5 requires all electrical cables and conductors to be listed for use in “wet locations” where installed underground, even if they are contained within a raceway or other enclosure (Fig. 300-16). This effectively applies to all raceways, cables, boxes, enclosures, etc., covered by Chap. 3.
Figure 300-17 shows the rule of part (C). Note that this rule has been clarified to avoid the former implication that the raceway always had to exit beyond the building perimeter. It is now clear that a raceway can go from one point to another below the floor grade of a building without needing to pass beyond the wall line. However, the requirement for all cables under a building to be in raceway still applies.
As shown at the top of Fig. 300-18, direct buried conductors or cables coming up a pole or on a building from underground installation must be protected from the minimum required burial distance below grade (from Table 300.5, but never required to be more than 18 in. [450 mm] into the ground) to at least 8 ft (2.5 m) above grade, as required by part (D) of this section. Where exposed to physical damage, raceways on buildings and raceways on poles must be rigid conduit, IMC, PVC Schedule 80, or equivalent, and the raceway or other enclosure for underground conductors must extend from below the ground line up to 8 ft (2.5 m) above finished grade. If a raceway on a building or on a pole is not subject to physical damage, EMT or Schedule 40 PVC may be used instead of other raceways.
Fig. 300-14. These applications are also covered in the table burial depths. (Sec. 300.5.)
Figure 300-19 shows the service lateral protection ribbon rule from 300.5(D)(3), along with other requirements including the ground movement accommodation rule in 300.5(J) and a warning on an important limitation on straight underground USE cable not colisted with a building wire designation.
Fig. 300-15. Reduced burial depth for low-voltage landscape lighting and lawn-sprinkler controls. (Sec. 300.5.)
Figures 300-20 and 300-21 show other rules of 300.5. Part (E) covers direct burial splices, which do not require boxes but do need to comply with 110.14(B), which requires the splicing method to be listed for this application. There are twist-on wire connectors listed for direct burial, and they are an appropriate solution for single-conductor type USE; however, do not use these connectors for conventional Type UF cable. The direct-burial listing on Type UF cable applies to the entire cable assembly and not to the individual conductors within the jacket. Therefore, if a splice is made with the jacket stripped back and these connectors made up, the resulting exposure of individual conductors within the UF cable assembly to direct soil burial violates the UF cable listing. The splicing method must provide an outer surface that is continuous from cable jacket to cable jacket.
Note that part (F) specifically requires that backfilled trenches must contain any necessary protection for raceways or cables buried in the trench. It specifies that sand or suitable running boards of wood or concrete or other protection must be afforded in those cases where backfill consists of heavy stones or sharp objects that otherwise would present the possibility of damage to the cable or raceway.
Fig. 300-16. Conductors run underground must be listed for “wet locations.” (Sec. 300.5.)
Fig. 300-17. Burial of cable in earth is not permitted under a building. (Sec. 300.5.)
Part (I) of this section requires that an underground circuit made up of single-conductor cables for direct burial must have all conductors of the circuit run in the same trench. That rule raises the question: When an underground direct burial circuit is made up of conductors in multiple, must all the conductors be installed in the same trench? And if they are, is derating required for more than three current-carrying conductors in a trench, just as it would be for more than three conductors in a single raceway? The answer to both questions is yes.
The wording of the rule in part (I) clearly indicates that all the conductors making up a direct burial circuit of single conductors in parallel must be run in the same trench and must be “in close proximity.” The wording of 300.5(I) also requires that all conductors of a circuit be run in the same raceway if a raceway is used (with building wire suitable for wet locations, such as THW or THWN). But Exception No. 1 permits parallel conductor makeup in multiple raceways, with each raceway containing all hot, grounded, and (if used) grounding conductors of the circuit. And Exception No. 2 to 300.5(I) recognizes the use of “isolated phase” installations, provided the rules for paralleling conductors—given in 310.4—and the rules for reducing the effects of inductive heating in adjacent metallic materials—as given in 300.20—are satisfied.
Fig. 300-18. Conductors from underground must be protected. (Sec. 300.5.)
When multiple-conductor makeup of a circuit is installed with all the parallel-circuit conductors in the same trench, it is necessary to observe the rule of 310.15(B)(2) of the NE Code, which states:
Where single conductors . . . are installed without maintaining spacing for a continuous length longer than 24 in. (600 mm) and are not installed in raceways, the allowable ampacity of each conductor shall be reduced as shown . . . ”
This means that the same deratings must be made as when more than three conductors are used in a single conduit, as explained in 310.15(B)(2)(a). Certainly, direct burial single conductors are covered by that requirement because Table 310.16 specifically covers direct burial conductors.
These Code rules often make for tricky and troublesome applications. For instance, as shown in Fig. 300-22, an underground circuit of Type USE insulated aluminum conductors might be used for a 3-wire, single-phase service to a multifamily dwelling. Because that is a residential service, 310.15(B)(6) might be considered to gain a higher-than-normal ampacity for the conductors; however, the rule has been rewritten to exclude multifamily applications and it only applies to applications where the conductors only supply a one-family
Fig. 300-19. Underground installations must accommodate ground movement. [Sec. 300.5(J).]
dwelling unit. Assuming that a 400-A conductor ampacity is indicated by the calculated demand load from Art. 220, each phase leg of the service feeder must have an ampacity of 400 A. Refer to the 75°C column in Table 310.16: A 250 kcmil THW aluminum has an ampacity of 205 A. Two such conductors per hot leg and two for the neutral would give the required 400-A capacity for the service. But how should the parallel circuit be run? All the circuit conductors must be run in close proximity in the same trench, as required by 300.5(I). That means all six USE conductors are in the same trench; and, because the neutrals do not count as current-carrying conductors, the derating of these conductors run without spacing must be to 80 percent of the 200-A ampacity, as required for four conductors in the table of Table 310.15(B)(2). With each 250 kcmil THW aluminum now derated to 161 A (0.8 × 205), the ampacity of each hot leg is only 322 A (2 × 161).
Fig. 300-20. Underground wiring must satisfy these requirements. (Sec. 300.5.)
Fig. 300-21. For direct burial underground conductors, a box must be used at splice points, with conductors brought up in sweep ells and the box properly grounded—unless listed materials are used to make directly buried splices in the conductors in accordance with Sec. 110.14(B). (Sec. 300.5.)
The rule requiring all conductors to be in the same trench makes the circuit of two 250 kcmil THW aluminum per leg inadequate. In referring to Table 310.16, it now becomes necessary to pick a larger size of THW aluminum, such that, when derated to 80 percent of their Table 310.16 value, the two of them will still provide the required 400-A rating. A 350-kcmil THW aluminum has a normal rating of 250 A. Derated to 80 percent (250 × 0.8), it has the needed ampacity of 200 A, so that two in parallel per hot leg and neutral will have the ampacity of 400 A. If the two parallel sets of conductors could have been run in separate trenches, the 250 kcmil THW aluminum conductors would have met the need in terms of ampacity. Obviously the conductors selected for actual direct burial must be Type USE, or combine that rating with THW.
Exception No. 1 of part (I) permits an underground circuit to be made up in parallel in two or more raceways, without need for derating. But, in such cases, each raceway must contain one of each of the phase legs, a neutral (if used), and an equipment grounding conductor (if used). With “A-B-C-N” in each raceway of a multiple group, that would be the same type of multiple-conduit-parallel-conductor makeup as required and commonly used for aboveground circuits.
Exception No. 2 permits “isolated-phase” makeup of underground circuits in multiple conduits—all phase A conductors in one conduit, all phase B conductors in a second conduit, all phase C conductors in a third conduit, all neutrals in a fourth conduit—with an equipment grounding conductor (or conductors, if needed) installed in a fifth conduit or installed in each of the three conduits carrying the phase conductors. However, that makeup is permitted only where the conduits are nonmetallic and are “in close proximity” to each other. (See 300.20.) Remember that the raceways must terminate in a manner that avoids inductive heating using methods in 300.20(B), such as entering below an open-bottom switchboard, or through a nonferrous window.
Fig. 300-22. Literal application of Code rules often imposes stiff requirements. (Sec. 300.5.)
Part (J) makes it mandatory that frost heave and settling be accommodated. In those areas of the country that experience cyclical freezes and thaws, concern about frost damage to buried raceways and cables has generally been addressed either by installation below the frost line or by the use of expansion fittings and direct buried cables with slack. This is done to ensure that the raceway remains intact and its contained circuits remain operational. The Code now mandates the implementation of those materials and methods that will ensure the underground raceways and cables are not damaged by frost heave. The fine-print note (FPN) immediately following part (J) identifies methods to accomplish the desired effect.
Figure 300-19 shows the rules in action. Note that this rule applies regardless of whether the run from grade to termination is enough to raise a concern about thermal expansion and contraction [300.7(B)]. It also applies regardless of the thermal expansion coefficient of the wiring method, because it is the ground that is moving, not the raceway changing length, and therefore it applies equally to metallic and nonmetallic raceways alike. There is considerable loss experience associated with steel raceways being forced vertically upward through concentric knockouts until the raceway shorted against a line bus.
In addition, do not make the mistake of assuming that this rule only applies in areas subject to frost. It applies to “earth movement.” This includes settling conditions that tend to drag raceways downward to the point of exposing the service conductors within to touch, or pulling downward on intact locknut connections to the point where enclosures are removed from a wall or severely deformed to the point that the covers no longer work properly. The usual way to address this rule is through the use of expansion fittings, which are certainly acceptable, but expensive in the case of metallic models. In some cases other approaches have worked, such as an elliptical hole where a conduit body entered a building on a small branch circuit.
This author once reviewed a medium-voltage application where the feeder emerged from grade and went up the side of a new industrial occupancy to the roof. The engineers were so concerned about the likelihood of settlement that they specified a special pull box with telescoping sides that would allow for the conductors to gently sweep up out of the ground, crest, and go downward almost to grade, and then rise again and enter the vertical raceways up the side of the building. The arrangement was designed to allow the grade to settle several feet without compromising the required bend radius on the Type MV cables.
In part (K) the Code mandates that underground cables and raceways are installed using “direct boring equipment” must be “approved” for such installation. “Approved” means acceptable to the inspector. In this case, wording in the manufacturer’s instructions indicating that such installation is permitted will generally satisfy the local inspector.
300.6. Protection Against Corrosion. This section provides some specific wording with regard to what must be provided to achieve compliance with the rule here. Ferrous-metal equipment is covered in part (A). An exception exempts stainless steel from the necessity to have protective coatings.
Part (A)(1) limits the use of ferrous-metal equipment protected solely by enamel to damp and dry indoor locations. Part (A)(2) allows organic coatings to be applied to metallic boxes or cabinets to prevent corrosion when used outdoors, in lieu of the standard “4-dip” zinc galvanizing method. And part (A)(3) addresses the use of ferrous metal equipment in concrete, earth, or other hostile environments, provided either the material or the corrosion protection is approved for the location.
Part (B) covers aluminum raceways, enclosures, etc. All such equipment must be provided with supplemental where in concrete or the earth. In the previous cycle, this part referred to nonferrous metal, which was an error because red brass is not subject to this limitation. Part (C) covers nonmetallic equipment, which must be suitable for use in direct sunlight and impervious to other environmental hazards based on its chemical makeup.
Figure 300-23 shows the right and wrong ways of installing equipment in indoor wet locations, as covered in part (D) of this section. Note that the exception that allows nonmetallic products to avoid the 6-mm (¼ in.) air space rule is limited to surfaces that are impervious to moisture, such as tile or concrete.
Fig. 300-23. Water or moisture must not be trapped in contact with metal. (Sec. 300.6.)
300.7. Raceways Exposed to Different Temperatures. Part (A) requires protection against moisture accumulation. If air is allowed to circulate from the warmer to the colder section of the raceway, moisture in the warm air will condense in the cold section of the raceway. This can usually be eliminated by sealing the race-way just outside the cold rooms so as to prevent the circulation of air. Sealing may be accomplished by stuffing a suitable compound in the end of the pipe (Fig. 300-24). Note that this rule has been expanded in recent years, and now expressly includes a passage from inside to outside a building (presumably only where there is a known issue, as with conditioned air). The former limitation to refrigerated rooms, as suggested in the sketch, is not the only mandatory application of this sealing rule.
The rule of part (B) makes it mandatory to ensure that expansion fittings are used where raceways are exposed to different temperatures to ensure that the raceways, fittings, and enclosures are not damaged and compromised by the inherent expansion and contraction that results from thermal cycling. The FPN directs the reader to data regarding the use of expansion fittings for PVC conduit and provides guidance and data that applies to the expansion characteristics of IMC, EMT, and rigid steel conduit. Do not confuse this with ground movement as covered in 300.5(J). They are two different rules and different conditions that mandate their application, although the usual solution for both problems is an expansion fitting.
Fig. 300-24. Sealing protects against moisture accumulation in raceway. (Sec. 300.7.)
300.8. Installation of Conductors with Other Systems. Any raceway or cable tray that contains electric conductors must not contain “any pipe, tube, or equal for steam, water, air, gas, drainage, or any service other than electrical.”
300.9. Raceways in Wet Locations Above Grade. This new (2008 NEC) rule clarifies once and for all that the inside of a raceway in a wet location is presumed to be wet, and the enclosed wiring must be suitable for wet locations.
300.10. Electrical Continuity of Metal Raceways and Enclosures. This is the basic rule requiring a permanent and continuous bonding together (i.e., connecting together) of all noncurrent-carrying metal parts of equipment enclosures—conduit, boxes, cabinets, enclosures, housings, frames of motors, and lighting fixtures—and connection of this interconnected system of enclosures to the system grounding electrode at the service or transformer (Fig. 300-25). The interconnection of all metal enclosures must be made to connect all metal to the grounding electrode and to provide a low-impedance path for fault-current flow along the enclosures to ensure operation of overcurrent devices that will open a circuit in the event of a fault. By opening a faulted circuit, the system prevents dangerous voltages from being present on equipment enclosures that could be touched by personnel, with consequent electric shock to such personnel.
Simply stated, this interconnection of all metal enclosures of electric wires and equipment prevents any potential aboveground on the enclosures. Such bonding together and grounding of all metal enclosures are required for both grounded electrical systems (those systems in which one of the circuit conductors is intentionally grounded) and ungrounded electrical systems (systems with none of the circuit wires intentionally grounded).
Exception No. 1 to this rule recognizes the specific permission given in Art. 250 to depart from the requirement for “electrical continuity,” where an isolated metallic elbow is used underground. As indicated in 250.86, Exception No. 2, short sections of metallic piping used for providing physical protection for cables are not subject to the basic requirement that such metallic raceways, cable jackets, etc., all be electrically continuous. But effective equipment interconnection and grounding are extremely important for grounded electrical systems to provide the automatic fault-clearing that is one of the important advantages of grounded electrical systems. A low-impedance path for fault current is necessary to permit enough current to flow to operate the fuses or CB protecting the circuit.
Fig. 300-25. All metal enclosures must be interconnected to form “a continuous electric conductor.” (Sec. 300.10.)
Exception No. 2 recognizes the permission given in 250.96(B) to interrupt the electrical continuity of a raceway containing a so-called isolated ground circuit. But, such installation must fully comply with all requirements given in 250.96(B) to ensure a modicum of safety.
300.11. Securing and Supporting. The basic rule in part (A) requires all race-ways, cables, boxes, and so on to be “securely fastened in place.” That wording effectively prohibits simply laying raceways, cables, boxes, and so on on the floor, on piping, on ceiling tiles, or anywhere. Although such installation may provide “support,” the basic rule calls for fastening, not just support. Regardless of where raceways, cables, boxes, and the like are installed, they must be fastened in place using a recognized method. Generally speaking, the rule in part (A) permits the use of “ceiling support” wires to secure wiring methods and equipment; however, they must be “in addition to the ceiling grid support wires.” That is, the ceiling will have a designed pattern of support wires, and tie wires for electrical equipment are to be in addition to the ones required to hold up the ceiling. This is accomplished by using appropriate commercially available fasteners. Note that where ceiling support wires are used, they must be secured at both ends. The last sentence simply reiterates a long-held interpretation that asserts that unfastened cables and raceways are prohibited, and that if tied to a support grid the adjacent ceiling panel will be very difficult or impossible to remove. Note that 300.23 specifically requires that any wiring above such panels be arranged so the panels can be removed and equipment accessed above the ceiling.
In part (A)(1), the Code calls for an independent means of support for wiring methods within the cavity above a suspended ceiling that is an integral part of fire-rated floor-ceiling or roof-ceiling assembly. This rule covers those installations where the fire rating of the hung ceiling is being used to establish the fire separation rating—typically specified by local building codes—for the space above and beneath the ceiling. Such installations are covered by (A)(1). Nonfire-rated hung-ceiling assemblies are covered under part (A)(2).
A true (A)(1) application is somewhat unusual, and it is important to know exactly what it is and why the rules make the distinctions they do. Building codes typically establish required fire separation ratings between floors of major buildings. Usually, the building is simply designed with enough concrete so the structural ceiling possesses the required fire separation by itself. In these cases the integrity of the suspended ceiling has no bearing on how much time it would take for a fire on one floor to compromise the next higher floor. These ceilings are covered in (A)(2).
However, there are other ways to achieve a particular floor/ceiling separation. Some ceilings are designed with a thinner slab than one that would establish the required fire separation on its own. In this case, the suspended ceiling is an integral part of the overall fire rating. Qualified testing laboratories have evaluated many such designs, using specified ceiling panel materials. These tests involve actually constructing such a room and ceiling inside a giant furnace, so they are extremely expensive to run. Invariably, the designs specify, in precise detail, exactly how many support wires must be used, how thick they must be, how and where they must be attached, how fixtures are to be supported, how air ducts must be constructed and run, etc.
In making these tests, the testing laboratories do not assume any additional weight loads on these support wires, nor do they (or could they) evaluate how such a ceiling might deform under fire conditions with such loading. In these cases, the wiring methods must be supported securely and independently of the support wires. The only exception allows for the remote possibility that someone might rerun a ceiling assembly test with specified wiring attached to specified supports. The fact that the rated designs specify the support wire locations makes it essential that both electrical and building inspectors be able to know which support wires were placed in addition to the essential elements of the ceiling construction.
Therefore, no raceways or cables may be supported by the wires installed to support the ceiling. But, if additional wires are installed, such wires may be used to “securely support” the raceways or cables within the ceiling. And because of the critical nature and precise construction detail that goes into such a ceiling design, where additional ceiling support wires are used as the support for wiring methods within a hung ceiling, such wires must be identified via marking, tagging, painting, or virtually any method that will permit the ceiling support wires to be readily distinguished from the electrical system support wires.
In part (A)(2) the Code addresses the same consideration for wiring method support within nonfire-rated hung ceilings. In the case of nonfire-rated applications, you have to add wires to those required for ceiling support, but you don’t have to identify them in any particular way. The language in this paragraph talks of an “independent means of secure support” which has led some to conclude that all support wires were off limits. That has never been the case and the issue was clarified in the 2008 NEC which added the phrase “and shall be permitted to be attached to the assembly.” Thus, as in the case of the (A)(1) applications, you can’t use the ceiling design wires but you can add to them, provided they are attached at both ends so as to “provide secure support” as covered in the parent language in 300.11(A). Of course, who is to say that the ceiling installer didn’t add a couple extra wires and those just happened to be the ones the electrician used. In the real world this is difficult to police.
Both (A)(1) and (A)(2) have rather impractical exceptions that recognize the use of the ceiling design wires if the design includes the wiring. The (A)(1) exception would require the entire ceiling system to go back in for testing, which is so costly as to unimaginable. Remember, the exact location of the wiring would need to be specified, and that is something that changes with every installation. The (A)(2) exception is predicated on the ceiling manufacturer choosing to recognize wiring on his specified ceiling support protocol. Frankly the likelihood of that happening is only slightly better than infinitesimal, but the exception is probably harmless as long as no one expects to use it.
In part (B), raceways are prohibited from being used as a means of support for cables or nonelectrical equipment. Telephone or other communication, signal, or control cables must not be fastened to electrical conduits, such as by plastic straps or any other means. Although raceways must not be used as a means of support for other raceways, cables, or nonelectric equipment, part (B)(1) might permit large conduits with hanger bars or fittings intended to support smaller raceways, but this application would have to be called out on some manufacturer’s literature, as provided in the definition of “identified.” Part (B)(2) permits such applications as tying Class 2 thermostat cable to a conduit carrying power-supply conductors for electrically controlled heating and air-conditioning equipment that is controlled by the Class 2 wires. Part (B)(3) applies to those instances where raceways are permitted to support boxes or conduit bodies (314.23) or fixtures [410.16(F)].
In part (C), the Code also prohibits the use of cable assemblies as a means of support for any equipment, such as other power cables, raceways, boxes, cable TV, or phone lines. In short, nothing may be supported by any of the cable assemblies recognized as wiring methods by the NEC. Two cables tie-wrapped together are a support for neither. The reason for the cable-on-other-wiring prohibitions is that the ampacity calculation procedures in the NEC presume that wiring methods are free to radiate heat. Adding a blanket of cables can lead to overheated wiring.
300.13. Mechanical and Electrical Continuity—Conductors. The general rule of part (A) is: Where conductors are spliced, a box is needed. The second sentence then identifies those limited number of cases where the Code does allow a splice without a box.
Part (B) prohibits dependency upon device terminals (such as internally connected screw terminals of duplex receptacles) for the splicing of neutral conductors in multiwire (3-wire or 4-wire) circuits. Grounded neutral wires must not depend on device connection (such as the break-off tab between duplex receptacle screw terminals) for continuity. White wires can be spliced together with a pigtail to the neutral terminal on the receptacle. If the receptacle is removed, the neutral will not be opened (Fig. 300-26).
Fig. 300-26. Neutrals of multiwire circuit must not be spliced at receptacle terminals. (Sec. 300.13.)
This rule is intended to prevent the establishment of unbalanced voltages should a neutral conductor be opened first when a receptacle or similar device is replaced on energized circuits. In such cases, the line-to-neutral connections downstream from this point (farther from the point of supply) could result in a considerably higher-than-normal voltage on one part of a multiwire circuit and damage to equipment, because of the “open” neutral, if the downstream line-to-neutral loads are appreciably unbalanced.
Note that this paragraph does not apply to 2-wire circuits or circuits that do not have a grounded conductor. This rule applies only where multiwire circuits feed receptacles or lampholders. This would most commonly be a 3-wire 240/120-V or a 3- or 4-wire 208/120-V or even a 480/277-V branch circuit.
The reason for the pigtailing requirement is to prevent the neutral conductor from being broken and creating downstream hazards. The problem lies in the inclination of electricians to work on hot circuits. Assume that a duplex receptacle on a 240/120-V 3-wire circuit becomes defective, and the first thing the electrician does, working hot, is to disconnect the neutral wires from the receptacle. Downstream, 2.4- and 12-A loads have been operating (plugged into additional receptacles on the multiwire circuit), each connected to a different hot leg. When the neutral is broken by the electrician upstream, normal operation of the loads reverts to the condition shown in Fig. 300-27. The two loads are now in series across 240 V. As shown, load A now has 200 V impressed across it. It could run extremely hot and burn out. Load B now has only 40 V across it; if it is a motor-operated device, the low voltage could cause the motor to burn up. Both could cause injuries. Also, in disconnecting the neutral, the electrician could get a 120-V shock if both the disconnected neutral conductor going downstream and the box were touched—not unlikely, since the neutral is usually considered to be dead—that is, at ground potential.
300.14. Length of Free Conductors at Outlets, Junctions, and Switch Points. The rule here applies only to the length of the conductor at its end. The exception covers wires running through the box. Wires looping through the box and intended for connection to outlets at the box need have only sufficient slack that any connections can be made easily. Generally, at least 150 mm (6 in.) of slack must be provided. This dimension is measured from the point the conductor emerges from its raceway or cable sheath. In addition, for smaller boxes (i.e., those smaller than 200 mm [8 in.] in any dimension), the minimum length of slack is based on the amount of conductor that extends outside the box—presumably when held at a 90° angle to its entry. For these smaller boxes, the conductor must be long enough so that at least 75 mm (3 in.) of conductor can extend from the box opening. This recognizes that boxes with small openings prevent a person from getting two hands into the box for the purpose of connecting devices, etc. Therefore, the Code mandates the conductors extend beyond the face of the box for not less than 75 mm (3 in.) where the box has an opening that is less than 200 mm (8 in.) in width or height.
Note the “emerges from its . . . cable sheath” provision. Those who don’t remove cable jackets prior to the walls being closed, and strip the jackets only at the finish wiring stage at the point where the cable assembly comes out of the wall may want to reconsider this practice. The 150-mm (6-in.) dimension would then start at the wall, resulting in excessive box fill. Although 314.16 has no prescriptive volume rule to cover this, it does require “free space for all enclosed conductors,” One of the reasons for specifying the point of emergence on cable assemblies as being the end of the jacket was to discourage this practice.
Fig. 300-27. Splicing neutrals on receptacle screws causes “open” in neutral if receptacle is removed. (Sec. 300.13.)
300.15. Boxes, Conduit Bodies, or Fittings—Where Required. The basic rule requires either a “box” or a “conduit body” to be used at splice points or connection points in cable and raceway systems. Figure 300-28 covers the basic principles. In addition, concealed knob-and-tube wiring must have a box at outlet and switch points. This sets the default stance of the NEC. The lettered paragraphs that follow, (A) through (M), set forth the only circumstances under which wiring methods can have any form of termination, pull, or outlet point without a box or conduit body.
Fig. 300-28. Raceways and cables may use boxes or conduit bodies at conductor splice points. (Sec. 300.15.) Note that based on current typical conduit body volumes, splices such as those shown are problematic because they usually will not meet the volume requirements in 314.16.
The second sentence in 300.15, also part of the default stance, was included to address the almost universal misuse of Type NM (Romex) connectors with other cables and even cords. The wording as it appears in the Code, however, requires that any fitting or connector be “designed and listed” for the wiring method used. Be aware that, while it will be up to the manufacturers of this equipment to obtain the listing for these products, it is the designer-installer’s responsibility to specify or use only those fittings and connectors specifically listed for the application. Connectors for Type MC cable are another source of abuse; many listed connectors for Type AC cable as misapplied on aluminum armored MC cable. The permitted cable types, including the allowable cable diameters and whether the cable is listed for smooth or corrugated or interlocking-armor cable or some combination of the three, will be listed on the smallest shipping unit carton.
In part (A) the Code eliminates the need for a box or conduit body where certain enclosures that have removable covers are used. And part (B) recognizes equipment that are provided with adequately sized junction box or wiring compartment, such as motor terminal housings, or wiring compartments in HID and other luminaires that are arranged for a direct wiring entry.
Part (C) calls for abrasion protection for cables where they enter/exit a conduit or tubing. This ensures that rough or sharp edges on the conduit or tubing will not damage the cable. This applies to kick pipes and also to short risers from outlet and switch points on unfinished basement walls. Part (D) exempts a specific Type MI splice, but only where the actual splice is accessible.
Part (E) of this section recognizes the use of wiring devices that have “integral enclosures.” These are the so-called boxless devices made and acceptable for use in nonmetallic sheathed cable systems (Type NM). Such listed devices [see 334.30(C) for the rules on these devices] do not require a separate box at each outlet because the construction of the device forms an integral box in itself. These devices are flush, modern versions of the old surface-mounted devices covered in (H). The devices in this latter paragraph are primarily used for surface installations of Type NM cable as covered in 334.40(B), although fished-in applications are recognized as well.
Part (F) recognizes transition from Type AC or MC cable to raceway without the need for a box, provided that no splice or termination is made in the conductors. This permits the common practice of changing from, say, AC to EMT for a run down a wall, with the armor stripped from a long length of the AC and the exposed wires run in the EMT. A suitable fitting made for connecting AC to EMT must be used (Fig. 300-29).
Part (G) correlates with the rules of 300.5(E) and 110.14(B), which permit splicing of underground direct burial cables and conductors using listed splice kits without a box.
Fig. 300-29. This type of no-box connection for cable-to-raceway change is permitted by part (F). [Sec. 300.15(F).]
Part (I) covers terminations and pull points in large accessible enclosures much larger than the wiring methods, where a box would serve no purpose. Examples include switchboards, motor control centers, and cabinets and cutout boxes generally.
Part (J) permits splices to be made within luminaire wiring compartments where the branch-circuit wires are spliced to fixture or ballast wires.
Part (K) correlates with provisions for radiant electric heating cable connections that are done with the connections embedded with the cables.
Part (L) covers handholes and manholes unless there is an actual connection to electrical equipment contained in the handhole or manhole, as opposed to a simple conductor or cable splice.
Part (M) is obsolete. It covers connections to one of the specially listed devices for use on closed-loop power distribution systems. The article authorizing these systems (Art. 780 in the 2005 NEC) has been withdrawn for lack of interest. Refer to the discussion at the end of Chap. 7 for more information on these systems.
300.16. Raceway or Cable to Open or Concealed Wiring. Where the wires are run in conduit, tubing, metal raceway, or armored cable and are brought out for connection to open wiring or concealed knob-and-tube work, a fitting such as that shown in Fig. 300-30 may be used. Where the terminal fitting is an accessible outlet box, the installation may be made as shown in Fig. 300-30.
300.18. Raceway Installations. This section presents a general Code concept requiring raceways to be installed as a complete system, with associated boxes and other enclosures, before pulling conductors into the raceway and box system. This rule is in the NEC because of reports of damage to conductors being pulled into incomplete raceway systems. There is another, more fundamental reason for this rule. Any raceway system is supposed to be capable of having its enclosed conductors withdrawn and replaced without damage (see 300.17). On very long runs, the frictional forces can overcome the tensile strength of the conductors and the ability of the insulation to stand up to the sidewall bearing pressures that will develop on difficult pulls. To avoid these problems, some contractors have pulled wire in segments, adding raceway lengths as they go forward, eventually resulting in a raceway that might as well be classified as a cable because there will be no possibility that the wires could ever be withdrawn. This section provides a partial defense against this by assuring that, at least, someone actually pulled the wires in through all the bends, from access point to access point.
The rule also intends to permit wiring of motors and fixture whips after the basic raceway system has been wired, as well as covering prewired assemblies. Figure 300-31 shows wires pulled into an incomplete raceway system. That used to be a Code violation, but the rule prohibiting it was removed as long as the raceways are complete between pull points. The raceway to the right is still OK because it has no wires in it. The Exception recognizes the permission given in 250.86 Exception No. 2. Part (B) prohibits welds made to raceways unless they were specifically designed for this purpose, or some Code provision specifically allows it.
Fig. 300-30. These are techniques for connecting conduit to open wiring. (Sec. 300.16.)
300.19. Supporting Conductors in Vertical Raceways. Long vertical runs of conductors should not be supported by the terminal to which they are connected. Supports, as shown in Fig. 300-32, may be used to comply with 300.19(A), as may the other support methods recognized by part (B). (See also Figs. 300-33 and 300-34.)
Fig. 300-31. Conductors shown here have been pulled into the conduit before boxes and continuation of the raceway system were installed to supply underground circuits to outdoor building lighting. Under previous Code editions, this was a violation of Sec. 300-18. That section and its rules were removed after the 1981 NEC, but restored six years later.
Fig. 300-32. Some type of support must carry the weight of conductors in long risers. (Sec. 300.19.)
Fig. 300-33. Bore-hole cable, with steel wire armor, is permitted by the exception to be supported only at the top of very high risers because the steel armor supports the length of the cable when the steel wires are properly clamped in the support ring of the type of fitting shown here. (Sec. 300.19.)
300.20. Induced Currents in Metal Enclosures or Metal Raceways. When all conductors of an AC circuit are kept close together—in a raceway or a box or other enclosure—the magnetic fields around the conductors tend to oppose or cancel each other, thereby minimizing the inductive reactance of the circuit and also minimizing the amount of magnetic flux that can cause heat due to hysteresis loss (magnetic friction) in steel or iron and due to the I2R losses of currents that are induced in adjacent metal. The rule of this section calls for always running a neutral conductor with the phase legs of an AC circuit to minimize such induction heating. The equipment grounding conductor must also be run close to the circuit conductors to achieve the reduction in inductive reactance and minimize the impedance of the fault-current return path when a fault does occur, thereby assuring the fastest possible operation of the protective device (fuse or CB) in the circuit. The reference to “all” grounding conductors in the basic rule is aimed at the so-called isolated grounding conductors, which must also be run with the circuit conductors from the outlet to the panel where that circuit originates. Where the isolated ground continues through that panel, it should be run with the feeder that supplies that panel, although that is not entirely clear (Fig. 300-35).
Fig. 300-34. Separate strands of cable armor are snubbed between flanges of support fitting at top of run. Partitioned enclosure protects unarmored sections of cable. (Sec. 300.19.)
When an AC circuit is arranged in such a way that the individual conductors are not physically close for mutual cancellation of their field flux, it is particularly important to take precautions where a single conductor passes through a hole in any magnetic material, such as a steel enclosure surface. The presence of the magnetic material forms a closed (circular) magnetic core that raises the flux density of the magnetic field around the conductor (i.e., it greatly strengthens the magnetic field). Under such conditions, there can be substantial heating in the enclosure due to hysteresis (friction produced by the alternating reversals of the magnetic domains in the steel) and due to currents induced in the steel by the strong magnetic field. To minimize those effects, the second paragraph of Sec. 300.20 requires special treatment, such as that shown in Fig. 300-36. Or a rigid, nonmetallic board (fiberglass, plastic, etc.) should be used for the enclosure wall that the conductors pass through.
Fig. 300-35. Close placement of AC conductors minimizes magnetic fields and induction. (Sec. 300.20.)
300.21. Spread of Fire or Products of Combustion. Application of this section to all kinds of building constructions is a very broad and expanding controversy in modern electrical work, in particular because of the phrase substantially increased. The rule here requires that electrical installations shall be made to substantially protect the integrity of rated fire walls, fire-resistant or fire-stopped walls, partitions, ceilings, and floors. Electrical installations must be so made that the possible spread of fire through hollow spaces, vertical shafts, and ventilating or air-handling ducts will be reduced to a minimum. These rules require close cooperation with building officials to avoid destruction of fire ratings when electrical installations extend through such areas.
Fig. 300-36. Induction heating is reduced by opening the magnetic core. (Sec. 300.20.)
Certainly, poke-through wiring—that technique in which floor outlets in commercial buildings are wired through holes in concrete slab floors—is an acceptable wiring method if use is made of UL-listed poke-through fittings that have been tested and found to preserve the fire rating of the concrete floor. Throughout the country, poke-through wiring continues to be a popular and very effective method of wiring floor outlets in office areas and other commercial and industrial locations. Holes are cut or drilled in concrete floors at the desired locations of floor outlets, and floor box assemblies are installed and wired from the ceiling space of the floor below. The method permits installation of every floor box at the precise location that best serves the layout of desks and other office equipment.
The wiring of each floor outlet at a poke-through location may be done basically in either of two ways—by some job-fabricated assembly of pipe nipples and boxes or by means of a manufactured through-floor assembly (Fig. 300-37) made expressly for the purpose and tested and listed by a nationally recognized testing lab, such as UL.
May either of the methods be used? A clear regulation of the Occupational Safety and Health Administration (OSHA) appears to rule decisively on this question. OSHA clearly and flatly demands that an installation or equipment determined to be safe by a nationally recognized testing lab must always be used in preference to any equipment not certified by a testing lab. Thus, if a UL-listed poke-through fitting is available, then the use of any nonlisted, homemade assembly that has not been determined to be safe appears to be clearly not acceptable to OSHA.
Fig. 300-37. Several manufacturers make UL-listed poke-through assemblies. (Sec. 300.21.)
300.21 also applies to cable and/or conduit penetrations of fire-rated walls, floors, or ceilings without altering the fire rating of the structural surface (Figs. 300-38 and 300-39).
Fig. 300-38. Fire-stop fitting for passing cables or conduit through a fire-rated wall, floor slab, or similar concrete surface, without altering the fire rating of the surface. (Sec. 300.21.)
Fig. 300-39. Another type of device to provide for passing cable and/or conduit through fire-rated building surfaces without altering the conditions of fire resistance. (Sec. 300.21.)
Another similar concern covered by this section is the installation of lay-in lighting fixtures in a fire-rated suspended ceiling. Suspended ceilings are usually evaluated only for their esthetic and acoustic value, but they also serve as fire-protective membranes for the floor above. Although concrete floor structures have various fire-resistance ratings by themselves (depending on the concrete thickness and aggregate used), some assemblies require some type of protective cover. When this is the case, the ceiling is tested in combination with the floor-slab structure for which the rating is desired. Such a ceiling properly serves its function of fire protection until an installer cuts holes in it, such as for recessed lighting fixtures or for air diffusers or grilles. Because of that, the acceptability of the overall ceiling system must be carefully determined. Refer to the discussion at 300.11(A)(1) for more information on this part of the process.
The FPN following the basic rule provides some guidance, but the following presents a comprehensive step-by-step procedure for using the publications vaguely referenced in the FPN. First, check the Underwriters Laboratories’ Electrical Construction Materials Directory (commonly called the UL Green Book), which notes that recessed fixtures that have been shown to provide a degree of fire resistance with the floor, roof, or ceiling assemblies with which they have been tested are labeled as follows: “Recessed-type electric fixture classified for fire resistance; fire-resistance classification floor and ceiling Design No.——.” Next, find the design referred to in the UL Fire Resistance Index. This booklet follows the format of the Electrical Construction Materials Directory. Refer to a design of the required fire rating and be sure that the fixtures are listed for use with that design.
Designers must specify the particular UL design that suits their requirement, note this in the specifications, and be certain that the lighting fixtures are fire-rated in accordance. But it is advisable for the electrical contractor to investigate the ceiling design for possible fire rating in all cases and to receive from the designer written confirmation of the exact nature and value of the rating, if one exists.
In the UL Building Materials Directory, various fire-rated assemblies are listed by design number and by rating time. A companion publication, the Fire Resistance Index, contains detailed cross-sectional drawings of the assemblies, with all critical dimensions shown. Each pertinent element is usually flagged with an identifying number. Keyed to the number are clarifying statements listing additional critical limitations (such as the size and number of penetrations in the ceiling).
The top installation in Fig. 300-40 was tested and given a 1½-h rating. No protective material was used between the fixture and the floor slab above. A somewhat better rating could have been obtained had protection been provided over the fixture.
At the bottom of Fig. 300-40 is a fixture with protection. When this construction was tested, failure occurred after 2 h 48 min, and it received a 2-h time rating. Even with this type of protection, the UL listing will limit the area occupied by fixtures to 25 percent of the total ceiling area. (But a coffered ceiling may contain 100 percent lighted vaulted modules.)
Fig. 300-40. The complete assembly of concrete slab plus fixture and ceiling gets a fire rating. (Sec. 300.21.)
Plasterboard (gypsum board) panels used so commonly for interior wall construction in modern buildings are fire-rated. UL and other labs make tests and assign fire ratings (in hours) to wall assemblies or constructions that make use of plasterboard. For instance, a wall made up of wood or metal studs with a single course of
-in. (15.9-mm) plasterboard on each side of the studs would be assigned a 1-h fire rating. A wall with two courses of ½-or-in. (12.7- or 15.9-mm) plasterboard on each side of the studs would be a 2-h wall (Fig. 300-41). The assigned fire ratings are based on the thickness and number of courses of plasterboard. And the fire rating is for the wall assembly without any penetrations into the wall.
Fig. 300-41. Wall assemblies using plasterboard are firerated by UL and others. (Sec. 300.21.)
Because of the fire rating assigned to the assembly, any wall so constructed is fire-rated. The wall may be between rooms or between a room and a corridor or stairwell. No distinction is made between an interior wall of an apartment and a wall that separates one apartment from another. All wall assemblies using plasterboard are fire-rated and immediately raise concern over violation of 300.21 if any electrical equipment is recessed in the wall. However, this is one area where an intelligent use of the phrase “substantially increased” may be useful. If the building code considers both sides of a given wall, such as the one between a dwelling unit bedroom and the hallway leading to it, as being in the same smoke compartment and not deserving of any particular fire rating by code, then it is difficult to imagine how using fire-rated electrical penetrations accomplishes any improvement in safety.
Building inspectors and electrical inspectors have generally permitted installation of wall switches, thermostats, dimmers, and receptacles in boxes recessed in plasterboard walls. In single-family houses, the entire interior is not considered to be compartmented. It is assumed that individual rooms or areas are not normally closed off from each other and that fire or smoke spread would not be affected at all by those penetrations. The consensus has been that such small openings cut in the plasterboard do not violate the letter or intent of 300.21.
In apartment houses, office buildings, and other multioccupancy buildings, however, inspectors could logically question use of wiring devices installed in common walls between apartments or between an apartment and a corridor or stairwell. Such walls are assumed to be between interior spaces that are normally closed off from each other by the main doors to the individual apartments. Fire and/or smoke spread, which are normally restricted by the closed doors, might be considered substantially increased by any penetrations of those fire-rated walls (Fig. 300-42). Although switch and receptacle boxes are usually accepted, use of a panelboard in a common wall has been rejected.
Fig. 300-42. Walls separating closed-off spaces must have maintained fire rating. (Sec. 300.21.)
Another issue frequently arises in mirror image construction in apartment houses, where one kitchen will back up to the kitchen in the next unit to save on plumbing costs. The default separation rule on boxes is 600 mm (24 in.) of horizontal separation, and that dimension applies even if framing members intervene. With boxes required every few feet in kitchen counters, these rules need careful attention. There are now listed nonmetallic boxes that do not require the full horizontal separation. In addition, there are now listed “putty pads” that can be used to wrap around metal boxes that also dramatically reduce the required separation.
For larger electrical equipment, such as panelboards, the same general analysis would apply. For interior walls of private houses or individual unit occupancies in apartment houses, hotels, dormitories, office buildings, and the like, a panel installed in a wall between two rooms or spaces that are normally not closed off from each other cannot “substantially” contribute to greater fire and/or smoke spread. However, panelboards and similar large equipment should normally not be installed in fire-rated walls between spaces that are closed off from each other by doors that are normally closed.
When it is necessary to install a panelboard or other large equipment in a wall between areas that are normally closed off from each other, a boxed recess in the wall should be constructed of the fire-rated plasterboard to maintain the fire rating of the wall (Fig. 300-43). This is also a common practice for installing recessed enclosures for fire extinguishers in corridor walls and medicine cabinets mounted in walls between apartment units.
Fig. 300-43. Boxing of large-area penetrations has been required in fire walls. (Sec. 300.21.)
Another technique that has been used to maintain fire rating where a panel-board is installed in a wall between individual apartments is to glue pieces of plasterboard to the top, bottom, sides, and back of the recessed panel. In one particular job, this was done as a corrective measure where panelboards had first been installed in such walls without attention to maintaining the fire rating of the wall. But the use of plasterboard directly affixed to the panelboard surfaces could be considered an unauthorized modification of the panel that voids UL listing because of improper application.
300.22. Wiring in Ducts, Plenums, and Other Air-Handling Spaces. Part (A) of this section applies only to wiring in the types of ducts described, which can contain no wiring systems of any type. Part (B) covers use of wiring methods and equipment within “ducts or plenums,” which are channels or chambers intended, specifically fabricated, and used only for supply or return of conditioned air. Such ducts or plenums are sheet metal or other types of enclosures that are provided expressly for air handling and must be distinguished from “Other Space Used for Environmental Air”—such as the space between a suspended ceiling and the floor slab above it. Space of that type is covered by part (C) of this section. The space between a raised floor (Fig. 300-44) and the slab below the raised floor is also covered by part (C), unless the air-handling raised floor is within a computer room. If the raised floor is in a computer room, as defined in Art. 645, part (D) states that such an air-handling raised floor used for data-processing circuits must comply with Art. 645.
Fig. 300-44. Space under a raised floor, which is commonly used for circuits to data-processing equipment and provides for passage of conditioned air to the room and to the equipment is covered in part (D). (Sec. 300.22.)
One NFPA Standard defines a duct system as “a continuous passageway for the transmission of air which, in addition to ducts, may include duct fittings, dampers, plenums, fans, and accessory air handling equipment.” The word duct is not defined, but a plenum is defined as “an air compartment or chamber to which one or more ducts are connected and which forms part of an air distribution system.”
Part (B) addresses actual duct systems with air movement and excludes all wiring with any nonmetallic coverings and with pervious enclosures. Thus, Type MC cable is acceptable, but only in its corrugated or smooth forms, and never with a nonmetallic jacket. Type MI cable is acceptable as well. Flexible metallic tubing, rigid and intermediate metal conduit, and electrical metallic tubing are permitted for the same reasons. Greenfield is permitted, but only where absolutely needed to repositionable equipment required to be there and never in lengths over 1.2 m (4 ft). Enclosed and gasketed luminaires are also permitted in association with other necessary equipment.
Part (C) covers other spaces used for environmental air movement, particularly plenum cavity ceiling spaces, where the space above a hung ceiling point is used as a central collection point for return air to a central air handler. These spaces present a less intensive, but comparable concern as actual ductwork carrying the same air, and the NEC rules are also comparable, but slightly less intense as well. There is an important exception at this point, generally used for Type NM cable, that excludes joist or stud cavities used for return air movement, as shown in Fig. 300-46 This exception, however, has two very specific limitations. First, it only applies to dwelling units. Second, it only applies if the wiring involved passes straight across the short dimension. That means no turns, no device boxes, etc.; just straight across.
This part permits use of totally enclosed, nonventilated, insulated busway in an air-handling ceiling space, provided it is a non–plug-in-type busway that cannot accommodate plug-in switches or breakers. This one specific busway wiring method was added for hung-ceiling space used for environmental air. The methods in part B can be used, except that interlocking-armor Type MC cable can be used as well (but still no nonmetallic covering is allowed). Surface metal raceway or wireway with metal covers or solid-bottom metal cable tray with solid metal covers may be used in air-handling ceiling space, provided that the raceway is accessible, such as above lift-out panels. Other factory-assembled cable assemblies (without nonmetallic jackets) for use with Art. 604 manufactured wiring systems are allowable as well. (See Fig. 300-47.) The air-handling space under a raised floor in a data-processing location is covered by Art. 645 on information technology equipment. Figure 300-45 shows wiring methods for use in an air-handling ceiling space.
The Code panel has made it clear that they generally oppose nonmetallic wiring methods in ducts and plenums and in air-handling ceilings, except for nonmetallic cable assemblies that are specifically listed for such use. It is also the intent of the Code that cables with an outer overall nonmetallic jacket should not be permitted in ducts or plenums. Although the jacket material—usually PVC—would not propagate a fire, it would contribute to the smoke and provide additional flammable material in the air duct. The last sentence in part (C) is intended to exclude from the requirements those areas that may be occupied by people. Hallways and habitable rooms are being used today as portions of air-return systems, and while they have air of a heating or cooling system passing through them, the prime purpose of these spaces is obviously not air handling.
Fig. 300-45. A joist space through which Type NM cable passes “perpendicular to the long dimension” of the space may be closed in to form a duct-like space for the cold-air return of a hot-air heating system—but only in a “dwelling unit.”
Fig. 300-46. Any wiring method other than these is a violation in air-handling space. (Sec. 300.22.)
Fig. 300-47. Modular wiring systems, as recognized by Art. 604, are permitted to be used in air-handling ceiling spaces.
Part (C)(2) permits use of nonmetallic equipment enclosures and wiring that are specifically UL-listed or classified for use in air-handling ceiling spaces. The basic condition that must be satisfied is that the wiring materials and other construction of the equipment must be suitable for the expected ambient temperature to which they will be subjected.
In effect, the rules in parts (B) and (C) exclude from use in all air-handling spaces any wiring that is not metal-jacketed or metal-enclosed, to minimize the creation of toxic fumes due to burning plastic under fire conditions. 800.53(A), and similar rules in Art. 810, 820, and 830 covering radio, cable TV, and power broadband networks, basically require telephone, intercom, and other communications circuits to be wired with Type CMP cable or other identified types installed in compliance with 300.22 when such circuits are used in ducts or plenums or air-handling ceilings. Wiring in air-handling space under raised floors in computer centers must use the wiring methods described in 645.5(D). Ventilation in the raised-floor space must be used only for the data-processing area and the data-processing equipment.
Application of that Code permission on use of equipment calls for substantial interpretation. The designer and/or installer must check carefully with equipment manufacturers and with inspection agencies to determine what is acceptable in the air-handling space above a suspended ceiling. Practice in the field varies widely on this rule, and Code interpretation has proved difficult.
The exception to part (C)(2) recognizes the installation of motors and control equipment in air-handling ducts where such equipment has been specifically approved for the purpose. Equipment of this type is UL-listed and may be found in the Electrical Appliance and Utilization Equipment List under the heading “Heating and Ventilating Equipment.”
300.34. Conductor Bending Radius. This section covers the acceptable final bending radius of medium-voltage conductors in enclosures. This section is important because it is referenced in other locations in the NEC. For example, if a medium-voltage cable enters the back of a pull box opposite to a removable cover, this section sets the minimum depth allowed for the pull box. The dimensions are based on the diameters of the cables being installed. Great care must be used at medium-voltage cable terminations in enclosures, which may only be wide enough for straight-in terminations. Many failures have occurred after cables were bent into loops prior to landing in enclosures not wide enough to allow such loops to be formed without overstressing the shielding.
300.37. Aboveground Wiring Methods. Conductors aboveground must be in rigid metal conduit, in intermediate metal conduit (IMC), in electrical metallic tubing, in rigid nonmetallic conduit, in cable trays, in a cable bus, in busways, in other identified raceways, or as open runs of metal-clad cable suitable for the use and purpose. (Fig. 300-48). Rigid nonmetallic conduit has been added to the list of raceways acceptable for running medium-voltage circuits aboveground—and the PVC conduit does not have to be encased in concrete.
Fig. 300-48. A variety of wiring methods may be used for aboveground high-voltage circuits. (Sec. 300.37.)
In locations accessible to qualified persons only, open runs of Type MV cable, bare conductors, and busbars may be used (490.32 and 490.33). In locations accessible to qualified persons only, there are no restrictions on the types of wiring that may be used. The types more commonly employed are open wiring on insulators with conductors either bare or insulated, and rigid metal conduit or nonmetallic rigid conduit containing lead-covered cable.
300.35Protection Against Inductive Heating. This rule echoes 300.20, but it is worth noting that the higher power and voltages involved in these applications have produced dramatic failures in comparatively short order, as in within days of energization.
300.40. Insulation Shielding. One of the basic decisions to make in selecting medium-voltage conductors is whether or not electrostatic insulation shielding is required on the cable. The basic requirements on electrostatic shielding of medium-voltage conductors are presented in 310.6 and Table 310.5 and are explained there.
This section sets forth very general rules on terminating shielded medium-voltage conductors. The metallic shielding or any other conducting or semiconducting static shielding components on shielded cable must be stripped back to a safe distance according to the circuit voltage, at all terminations of the shielding. At such points, stress reduction must be provided by such methods as the use of potheads, terminators, stress cones, or similar devices. The wording of this regulation makes it clear that the need for shield termination using stress cones or similar terminating devices applies to semiconducting insulation shielding as well as to metallic-wire insulation shielding systems.
A stress cone is a field-installed device or a field-assembled buildup of insulating tape and shielding braid that must be made at a terminal of medium-voltage shielded cable, whether a pothead is used or not. A stress-relief cone is required to relieve the electrical stress concentration in cable insulation directly under the end of cable shielding. Some cable constructions contain stress-control components that afford the cable sufficient stress relief without the need for stress-relief cones. If a cable contains inherent stress-relief components in its construction, that would satisfy 300.40 as doing the work of a stress cone. As a result, separate stress cones would not have to be installed at the ends of such cable. Or, heat-shrinkable tubing terminations may be used with stress-control material that provides the needed relief of electrostatic stress.
At a cable terminal, the shielding must be cut back some distance from the end of the conductor to prevent any arcing over from the hot conductor to the grounded shield. When the shield is cut back, a stress is produced in the insulation. Providing a flare-out of the shield, that is, extending the shield a short distance in the shape of a cone, relieves the stress, as shown in Fig. 300-49.
Stress cones provide that protection against insulation failure at the terminals of shielded medium-voltage cables. Manufacturers provide special preformed stress cones (Fig. 300-50) and kits for preparing cable terminals with stress cones for cables operating at specified levels of medium voltage (Fig. 300-51). A wide assortment of stress-relief terminators are made for all the medium-voltage cable assemblies used today.
Fig. 300-49. Here is how a stress cone protects insulation at cable ends. (Sec. 300.40.)
Metallic shielding tape must be grounded, as required by 300.40 and 300.5(B), which refer to “metallic sheathing,” as in Fig. 300-52. The shield on shielded cables must be grounded at one end at least. It is better to ground the shield at two or more points. Grounding of the shield at all terminals and splices will keep the entire length of the shield at about ground potential for the safest, most effective operation of the cable. Cable with improperly or ineffectively grounded shielding can present more hazards than unshielded cable.
In addition to the requirements at terminations that apply to shielding, there is a related topic, that of the actual connector at the end, such as a lug. It should be designed to be compatible with medium-voltage applications. Its external surfaces should be well rounded to minimize the production of corona discharge. These lugs are available as listed items.
300.42. Moisture or Mechanical Protection for Metal-Sheathed Cables. A pothead is one specific form of stress-reduction means referred to in 300.40 and has long been a common means of protecting insulation against moisture or mechanical injury where conductors emerge from a metal sheath (Fig. 300-53). Such protection for metal-sheathed cables (such as lead-covered, paper-insulated cables) is required by this section. A pothead is a cable terminal that provides sealing to the sheath of the cable for making a moisture-proof connection between the wires within the cable and those outside.
When metal-jacketed medium-voltage cables are terminated outdoors exposed to the weather, a pothead is commonly used to protect the insulation of conductors against moisture or mechanical injury where conductors emerge from a metal sheath, as shown in Fig. 300-54.
Fig. 300-50. Typical preformed stress cone is readily applied on cables up to 35-kV indoors. (Sec. 300.40.)
On use of potheads:
1. Paper-insulated cables must be terminated in potheads. This requirement also extends to such cables operated at under 600 V.
2. Varnished-cambric-insulated cables should be terminated in potheads but may be terminated with taped connections in dry locations.
Fig. 300-51. “Pennant” method is one of a variety of job-site termination buildups. (Sec. 300.40.)
Fig. 300-52. Metallic shielding must be grounded for all high-voltage conductors—under- or aboveground. (Sec. 300.40.)
Fig. 300-53. Typical single-conductor pothead protects metallic- or nonmetallic-jacketed cable. (Sec. 300.42.)
3. Rubber-insulated cables are commonly terminated in potheads in locations where moisture protection is critical but may be terminated without potheads in accordance with manufacturer’s instructions.
4. Although many modern medium-voltage cables can be terminated without potheads, many engineers consider potheads the best terminations for medium-voltage cable.
5. The use of potheads offers a number of advantages:
a. Seals cable ends against moisture that would damage the insulation
b. Provides a compartment for surrounding the termination with insulating compound to increase the strength of electrical insulation
c. Seals cable ends against loss of insulating oils
d. Provides engineered support of connections
In the past, medium-voltage circuits used for commercial and industrial feeders, both outdoors and indoors, were covered in Art. 710. The 1999 Code eliminated Art. 710 and dispersed the rules formerly given in that article throughout the Code. Rules on wiring method are now contained in Part II of Art. 300. Although voltages up to 15,000 V (15 kV) are most commonly used for these circuits today, higher voltages (26 and 35 kV) are often used because they can offer economy for extremely large installations. Typical circuits today operate at 4160/2400 and 13,800/8000 V—both 3-phase, 4-wire wye hookups.
Fig. 300-54. A pothead is used on the end of each paper-insulated, lead-covered cable to protect against entry of moisture, with a wiped lead joint at the terminal. (Sec. 300.42.)
Modern medium-voltage circuits for buildings include overhead bare or covered conductors installed with space between the conductors, which are supported by insulators at the top of wood poles or metal tower structures; overhead aerial cable assemblies of insulated conductors entwined together, supported on building walls or on poles or metal structures; insulated conductors installed in metal or nonmetallic conduits or ducts run underground, either directly buried in the earth or encased in a concrete envelope under the ground; insulated conductors in conduit run within buildings; and multiple-conductor cable assemblies (such as nonmetallic jacketed cables, lead-sheathed cable, or interlocked armor cable) installed in conduit or on cable racks or trays or other types of supports. Another wiring method gaining wide acceptance consists of plastic conduit containing factory-installed conductors, affording a readily used direct earth burial cable assembly for underground circuits but still permitting removal of the cable for repair or replacement.
300.50. Underground Installations. Directly buried nonmetallic conduit-carrying medium-voltage conductors does not have to be concrete-encased if it is a type approved for use without concrete encasement. If concrete encasement is required, it will be indicated on the UL label and in the listing. 300.50 permits direct burial rigid nonmetallic conduit (without concrete encasement) for medium-voltage circuits.
Part (B) requires protection of conductors where they emerge from underground (see Fig. 300-55).
Fig. 300-55. Direct burial cables must be protected aboveground. (Sec. 300.50.)
300.50 makes it clear that underground circuits may be installed in rigid metal conduit, in intermediate metal conduit, or in rigid nonmetallic conduit. Rigid metal conduit or IMC does not have to be concrete-encased, but it may be, of course. Direct burial nonmetallic conduit must be an approved (UL-listed and labeled) type specifically recognized for use without concrete encasement. If rigid nonmetallic conduit is approved for use only with concrete encasement, at least 2 in. (50 mm) of concrete must enclose the conduit. All applications of the various types of nonmetallic conduit must conform to the data made available by UL in the Electrical Construction Materials Directory.
Special Condition (4) eliminates the need for any burial depth in earth for conduits or other raceways that are run under a building or exterior concrete slab not less than 4 in. (100 mm) thick and extending at least 6 in. (150 mm) “beyond the underground installation”—that is, overlapping the raceway by 6 in. (150 mm) on each side. Note that the 4-in. (100-mm)-thick concrete may be up at grade level in the form of a slab or patio or similar concrete area not subject to vehicular traffic. Figure 300-56 covers burial of underground medium-voltage circuits.
Fig. 300-56. Direct burial high-voltage cables must be of correct type, at specified depth. (Sec. 300.50.)
Part (A)(2) in 300.50 notes that other unshielded cable (i.e., cable without electrostatic shielding on the insulation)—other than within a metallic-sheathed cable assembly—must be installed in rigid metal conduit, in IMC, or in rigid nonmetallic conduit encased in not less than 3 in. (75 mm) of concrete. The effect of this rule is that unshielded, or nonshielded, cables may not be used directly buried in the earth. By reference to 310.7, nonshielded cables may be directly buried up to a rating of 2000 V, except that metal-encased, non-shielded conductors (as in Type MC or lead-jacketed cables) may be used in ratings up to 2400 V (other unshielded cables allowed up to 5000 V for series lighting at airfields). But all direct burial cables must be identified for such use. As indicated above, when nonshielded cable (of the nonmetallic-jacket type) is used underground in rigid nonmetallic conduit, the conduit must have a 3-in. (75-mm)-thick concrete encasement. But if the same nonshielded cable is used in rigid nonmetallic conduit aboveground, concrete encasement is not required.
Figure 300-56 demonstrates uses of direct burial medium-voltage cables in relation to the Code rules. Figure 300-57 covers underground medium-voltage circuits in raceways. Figure 300-58 covers the basic rules of Table 300.50, subject to the considerations noted in the rules and exceptions, as follows:
1. Where run under a building or slab that is at least 4 in. (102 mm) thick and that extends 6 in. (150 mm) in each direction beyond the cable or raceway.
2. Areas of heavy traffic (public roads, commercial parking areas, etc.) must have minimum burial depth of 24 in. (600 mm) for any wiring method.
3. In industrial locations only, with qualified maintenance and supervision, the minimum cover requirements for other than steel conduits can be reduced by 150 mm (6 in.) for each 50 mm (2 in.) of concrete or equivalent placed entirely within the trench over the underground installation. This applies to the general condition columns (1), (2), and (3). It does not modify the specific condition columns, which apply if applicable.
Fig. 300-57. Underground raceway circuits may vary widely in acceptable conditions of use. (Sec. 300.50.)
4. Lesser depths are permitted where wiring rises for termination.
5. Airport runways may have cables buried not less than 18 in. (450 mm) deep, without raceway or concrete encasement.
6. Conduits installed in solid rock may be buried at lesser depths than shown in diagrams when covered by at least 2 in. (50 mm) of concrete that extends to the rock surface.
Item 3 directly corresponds to Table 300.5, row 2 (for 600 V and below), with two differences. First, Table 300.50 allows the 50 mm (2 in.) provision to be duplicated, with successive reductions in burial for each thickness of concrete applied. This is appropriate because Table 300.50 has some required burial depths that are much lower than Table 300.5. In addition, Table 300.50 limits the permission to industrial occupancies, which is not done in Table 300.5. This permission (for medium voltage) had been in the NEC since 1968 but it dropped out when the table format was created, without substantiation. It has now returned, but with the industrial limitation, a limitation that was never included in any prior version. This may be problematic in the many institutional occupancies that operate medium voltage systems in campus layouts with similar qualified staffing.
Fig. 300-58. Underground high-voltage circuits must observe burial depths. (Sec. 300.50.) EMT, for the reasons given in the discussion at 300.5, is not a good choice for direct burial.
As noted in part (C), splices or taps are permitted in a trench without a box, but only if approved methods and materials are used—that is, listed splice kits. Taps and splices must be watertight and protected from mechanical injury. For shielded cables, the shielding must be continuous across the splice or tap. There is an exception to this continuity rule for engineered systems using direct-buried single-conductor cables with maintained spacing between phases. Under these conditions, the shielding is permitted to be made discontinuous, provided the severed ends of the shielding are overlapped at the point of discontinuity. Each shielded section is then separately grounded at a single point for each end. This interrupts the circulating currents from capacitive charging that become significant on long runs under these conditions. The currents can be high enough such that the shield heating seriously degrades the cable ampacity.
Figure 300-59 shows a permanent straight splice for joining one end of a cable off a reel to the start of a cable off another reel, or for repairing a cable that is cut through accidentally by a backhoe or other tool digging into the ground. T and Y splices are made with similar techniques. Disconnectable splice devices provide watertight plug-and-receptacle assembly for all types of shielded cables and are fully submersible.
Figure 300-60 covers the rules of parts (C) and (D).
Fig. 300-59. Splice may be made in direct burial cable if suitable materials are used. (Sec. 300.50.)
310.2. Conductors. Although conductors are generally required by this rule to be insulated for the phase-to-phase voltage between any pair of conductors, bare conductors may be used for equipment grounding conductors, for bonding jumpers, for grounding electrode conductors generally, and for grounded neutral conductors in certain locations (230.22, 230.30, 230.41, 230.62, 230.134, 230.140, and 338.3).
Fig. 300-60. Circuits must be protected and sealed where they enter equipment. (Sec. 300.50.)
The application shown in Fig. 310-1 is a commonly encountered violation of 310.2 because it involves an unauthorized use of a bare conductor. 250.140 Exception permits grounding of existing ranges, cooktops, and ovens to the neutral conductor, and then only where “the grounded conductor (the neutral) is insulated” or is a bare neutral of an SE cable that originates at the service equipment. Where supplied from a new circuit, a separate equipment ground must be provided.
310.2states that “conductors shall be insulated,” except where covered or bare conductors (see definition in Art. 100) are specifically permitted elsewhere in the Code. As noted above, the applicable rule in Art. 250 does not allow this arrangement on new installation, or on uninsulated neutrals under any other circumstance than Type SE cable is used, and only if the branch circuit originates at the service panel. The same objection would apply to any other wiring method.
Although the basic rule of this section requires conductors to be insulated, a note refers to 250.184 on the use of solidly grounded neutral conductors in medium-voltage systems. As an exception to the general rule that conductors must be insulated, 250.184 does permit a neutral conductor of a solidly grounded wye system to have insulation rated at only 600 V (Fig. 310-2). It also points out that a bare copper neutral may be used for service-entrance conductors or for direct buried feeders, and bare copper or copper-clad aluminum may be used for overhead sections of outdoor circuits.
Fig. 310-1. This is a controversial application that violates Secs. 310.2 and 334.108. (Sec. 310.2.)
Fig. 310-2. A note refers to neutral conductors of solidly grounded high-voltage systems. (Secs. 250.184 and 310.2.)
310.3. Stranded Conductors. Number 8 AWG and larger conductors must be stranded when they are installed in conduit, EMT, or any other “raceway.” The use of an insulated or stranded 8 AWG copper conductor is permitted as the equipment bonding conductor required by 680.23(B)(2)(b). But only a solid 8-AWG copper conductor is permitted by 680.26(B) at swimming pools for bonding together noncurrent-carrying metal parts of pool equipment—metal ladders, diving board stands, pump motor frames, lighting fixtures in wet niches, and so on. The reason is that in the first instance the wire will be pulled into a race-way, and in the second the conductor will be used for multiple bonding connections in exposed or corrosive locations where stranded wire does not survive.
310.4. Conductors in Parallel. The requirements of 310.4 for conductors in parallel recognize copper, copper-clad aluminum, and aluminum conductors in sizes 1/0 AWG and larger. Also, this section makes it clear that the rules for paralleling conductors apply to grounding conductors (except for sizing, which is accomplished in accordance with 250.122) when they are used with conductors in multiple.
Conductors that are permitted to be used in parallel (in multiple) include “phase” conductors, “polarity” conductors, “neutral” conductors, and “grounded circuit” conductors. In the places where this section describes parallel makeup of circuits, a grounded circuit conductor is identified along with phase, polarity, and neutral conductors to extend the same permission for paralleling to grounded legs of corner-grounded delta systems.
This section recognizes the use of conductors in sizes 1/0 and larger for use in parallel under the conditions stated, to allow a practical means of installing large-capacity feeders and services. Paralleling of conductors relies on a number of factors to ensure equal division of current, and thus all these factors must be satisfied in order to ensure that none of the individual conductors will become overloaded.
When conductors are used in parallel, all the conductors making up each phase, polarity, neutral, or grounded circuit conductor must satisfy the five conditions of part (B) in this section. Those characteristics—same length, same conductor material (copper or aluminum), same size, same insulation, and same terminating device—apply only to the paralleled conductors making up each phase, polarity, or neutral of a parallel-makeup circuit. All the conductors of any phase, polarity, or the neutral must satisfy the rule, but phase A conductors (all of which must be the same length, same size, etc.) may be different in length, material, size, etc., from the conductors making up phase B or phase C or the neutral. All phase B conductors must be the same length, same size, and so on; phase C conductors must all be the same; and neutral conductors must all be alike (Fig. 310-3). It is not the intent of the Code rule to require that conductors of one phase be the same as those of another phase or of the neutral. The only concern for safe operation of a parallel-makeup circuit is that all the conductors in parallel per phase leg (neutral or grounded conductor) will evenly divide the load current and thereby prevent overloading of any one of the conductors. Of course, the realities of material purchase and application and good design practice will dictate that all the conductors of all phases and neutral will use the same conductor material, will have the same insulation, will have as nearly the same length as possible to prevent voltage drop from causing objectionable voltage unbalance on the phases, and will be terminated in the same way. The size of conductors may vary from phase to phase or in the neutral, depending upon load currents.
Figure 310-4 shows two examples of parallel-conductor circuit makeup. The photo at bottom shows six conductors used per phase and neutral to obtain 2000-A capacity per phase, which simply could not be done without parallel conductors per phase leg. Note that a fusible limiter lug is used to terminate each individual conductor. Although limiter lugs are required by the NEC only as used in 450.6(A)(3), they may be used to protect each conductor of any parallel circuit against current in excess of the ampacity of the particular size of conductor. In addition to protecting against unequal division of current, limiters placed at both ends also increase overall reliability, because in the event of a ground fault on one of the conductors, the limiters will immediately de-energize only the faulted cable, often allowing the circuit to remain energized at reduced load until a complete shutdown can be arranged. The CB or fuses on such circuits are rated much higher than the ampacity of each individual conductor.
Fig. 310-3. This is the basic rule on conductors used for parallel circuit makeup. (Sec. 310.4.)
Fig. 310-4. These are examples of circuit makeup using conductors in parallel. (Sec. 310.4.)
Where large currents are involved, it is particularly important that the separate phase conductors be located close together to avoid excessive voltage drop and ensure equal division of current. It is also essential that each phase and the neutral, and grounding wires, if any, be run in each conduit even where the conduit is of nonmetallic material.
Part (C) of the section requires the use of the same type of raceway or enclosure for conductors in parallel. The impedance of the circuit in a nonferrous raceway will be different from the same circuit in a ferrous raceway or enclosure. (See 300.20.) In addition, each raceway must have the same number of conductors per phase. This is a recent addition to the NEC, after it was substantiated that someone actually ran three sets of parallel conductors through two raceways—two sets in one and one set in the other. Since the heat developed in the raceway with two sets of conductors will significantly differ from the other, and since the resistance of a conductor differs with temperature, the current will divide unequally.
From the Code tables of current-carrying capacities of various sizes of conductors, it can be seen that small conductor sizes carry more current per circular mil of cross section than do large conductors. This results from rating conductor capacity according to temperature rise. The larger a cable, the less the radiating surface per circular mil of cross section. Loss due to “skin effect” (apparent higher resistance of conductors to alternating current than to direct current) is also higher in the larger conductor sizes. And larger conductors cost more per ampere than smaller conductors.
All the foregoing factors point to the advisability of using a number of smaller conductors in multiple to get a particular carrying capacity, rather than using a single conductor of that capacity. In many cases, multiple conductors for feeders provide distinct operating advantages and are more economical than the equivalent-capacity single-conductor makeup of a feeder. However, it should be noted that the reduced overall cross section of conductor resulting from multiple conductors instead of a single conductor per leg produces higher resistance and greater voltage drop than the same length as a single conductor per leg. Voltage drop may be a limitation.
Figure 310-5 shows a typical application of copper conductors in multiple, with the advantages of such use. Where more than three conductors are installed in a single conduit, the ampacity of each conductor must be derated from the ampacity value shown in NEC Table 310.16. The four circuit makeups show:
1. Without ampacity derating because there are more than three conductors in the conduit, circuit 2 would be equivalent to circuit 1.
2. A circuit of six 400 kcmils can be made equivalent in ampacity to a circuit of three 2000 kcmils by dividing the 400s between two conduits [3 conductors/3-in. (76-mm) rigid metal conduit]. If three different phases are used in each of two 3-in. (76-mm) conduits for this circuit, the multiple circuit would not require ampacity derating to 80 percent, and its 670-A rating would exceed the 665-A rating of circuit 1.
3. Circuit 2 is almost equivalent to circuit 3 in ampacity.
4. Circuit 4 is equivalent to circuit 1 in ampacity, but uses less conductor copper and a smaller conduit. The advantages are obtained even with the ampacity derating for conduit fill.
Fig. 310-5. These circuit makeups represent typical considerations in the application of multiple-conductor circuits. (Sec. 310.4.)
Except where the conductor size is governed by conditions of voltage drop, it is seldom economical to use conductors of sizes larger than 1000 kcmil, because above this size the increase in ampacity is very small in proportion to the increase in the size of the conductor. Thus, for a 50 percent increase in the conductor size, that is, from 1,000,000 to 1,500,000 cmil, the ampacity of a Type THW conductor increases only 80 A, or less than 15 percent, and for an increase in size from 1,000,000 to 2,000,000 cmil, a 100 percent increase, the ampacity increases only 120 A, or about 20 percent. In any case where single conductors larger than 500,000 cmil would be required, it is worthwhile to compute the total installation cost using single conductors and the cost using two (or more) conductors in parallel.
Figure 310-6 shows an interesting application of parallel conductors. A 1200-A riser is made up of three conduits, each carrying three phases and a neutral. At the basement switchboard, the 1200-A circuit of three conductors per phase plus three conductors for the neutral originates in a bolted-pressure switch with a 1200-A fuse in each of the three phase poles. Because the total of 12 conductors make up a single 3-phase, 4-wire circuit, a 400-A, 3-phase, 4-wire tap-off must tap all the conductors in the junction box at top. That is, the three phase A legs (one from each conduit) must be skinned and bugged together and then the phase A tap made from that common point to one of the lugs on the 400-A CB. Phases B and C must be treated the same way—as well as the neutral. The method shown in the photo was selected by the installer on the basis that the conductors in the right-hand conduit are tapped on this floor, the center-conduit conductors are tapped to a 400-A CB on the floor above, and the left-conduit conductors are tapped to a 400-A CB on the floor above that. But such a hookup can produce excessive current on some of the 500 kcmils. Because it does not have the parallel conductors of equal length at points of load-tap, the currents will not divide equally, and this is a violation of the second paragraph of 310.4, which calls for parallel conductors to “be the same length.”
Fig. 310-6. A 1200-A circuit of three sets of four 500-kcmil conductors (top) is tapped by a single set of 500-kcmil to a 400-A CB (bottom) that feeds an adjacent meter center in an apartment house. This was ruled a violation because the tap must be made from all the conductors of the 1200-A circuit. (Note: The conduits feeding the splice box at top are behind the CB enclosure at bottom.) (Sec. 310.4.)
This arrangement is also an unauthorized tap, in violation of 240.21 because it amounts to creating a (in this case) one-third tap, and then (depending on conductor sizing) either using that tap at an excessive length, or tapping that tap, both of which violate the NEC. Review the discussion associated with Fig. 240-23 in Chap. 2 for more discussion of this concept.
Exception No. 1 of 310.4 permits parallel-circuit makeup using conductors smaller than 1/0 AWG—but all the conditions given must be observed.
This exception permits use of smaller conductors in parallel for circuit applications where it is necessary to reduce conductor capacitance effect or to reduce voltage drop over long circuit runs. It also recognizes the benefits in terms of lower reactance for high-frequency circuits (360 Hz and higher) as are commonly used in the aerospace industry. As it was argued in the proposal for this exception:
If a 14 AWG conductor, for example, is adequate to carry some load of not more than the 15-amp rating of the wire, there can be no reduction in safety by using two 14 AWG wires per circuit leg to reduce voltage drop to acceptable limits—with a 15-amp fuse or CB pole protecting each pair of 14 AWGs making up each leg of the circuit.
Where conductors are used in parallel in accordance with this exception, the rule requires that all the conductors be installed in the same raceway or cable. And that will dictate application of the last sentence of 310.4: “Conductors installed in parallel shall comply with the provisions of 310.15(B)(2)a” which applies to wire of 0 to 2000 V. Thus a single-phase, 2-wire control circuit made up of two 14 AWGs for each of the two legs of the circuit would have to be considered as four conductors in a conduit, and the “ampacity” of each 14 AWG would be reduced to 80 percent of the value shown in Table 310.16. If TW wires are used for the circuit described, the ampacity of each is no longer the value of 20 A, as shown in Table 310.16. With four of them in a conduit, each would have an ampacity of 0.8 × 20, or 16 A. Then using a 15-A fuse or CB pole for each pair of 14 AWGs would properly protect the conductors and would also comply with the 240.4(D)(3) which says that No. 14 must not have overcurrent protection greater than 15 A. (See Fig. 310-7.)
Exception No. 2 now recognizes the use of parallel conductors in sizes down to No. 2 where used as a neutral, but only in an existing installation. This is a good idea where necessary to accommodate additive harmonics on the neutral of multiwire circuits. The use of two No. 2s provides about 25 percent less cross-sectional area while at the same time providing 25 percent more surface area than a 3/0. This serves to reduce the heating caused by skin effect because the “skin” area has been increased.
Fig. 310-7. Overcurrent protection must be rated not in excess of the ampacity of one conductor when conductors smaller than No. 1/0 are used in parallel. (Sec. 310.4.)
The limitation to “existing installation” is somewhat puzzling since this technique would be useful in new construction. It seems as if the need for paralleling must first be demonstrated through measurements. In any event, paralleling grounded conductors smaller than 1/0, but not smaller than No. 2, may only be done at existing installations.
310.6. Shielding. The effect of this Code rule is to require all conductors operating over 2 kV to be shielded, unless the conductor is UL-listed for operation unshielded at voltages above 2 kV. The provisions that formerly allowed unshielded cables up to 8 kV have been deleted, replaced only by the 2400 V limit. Because 2300-V delta (which is over 2 kV) is the lowest general-purpose, medium-voltage circuit in use today, unlisted conductors must be shielded for such circuits. But note this: UL does list 2.4-kV unshielded conductors for use in accordance with Sec. 310.6, Table 310.13(D), and other Code rules (Fig. 310-8).
UL also lists shielded conductors up to 35 kV. And, in accordance with NE Code Table 310.13(B), UL has been listing Type RHH insulated conductors (rubber or cross-linked polyethylene insulation) with electrostatic shielding for operation up to 2 kV.
Fig. 310-8. A nonshielded conductor (arrow) is permitted for use on a 2300-V circuit (phase-to-neutral), as shown here, only if the conductor is listed by UL or another national test lab and approved for use without electrostatic shielding. (Sec. 310.6.)
In addition to applicable NE Code, Insulated Power Cable Engineers Association (IPCEA), and UL data on use of cable shielding, manufacturers’ data should be consulted to determine the need for shielding on the various types and constructions of available cables.
Shielding of medium-voltage cables protects the conductor assembly against surface discharge or burning (due to corona discharge in ionized air), which can be destructive to the insulation and jacketing. It does this by confining and distributing stress in the insulation and eliminating charging current drain to intermittent grounds. It also prevents ionization of any tiny air spaces at the surface of the insulation by confining electrical stress to the insulation. Shielding, which is required by this Code rule to be effectively grounded, increases safety by eliminating the shock hazard presented by the external surface of unshielded cables. By preventing electrical discharges from cable surfaces to ground, shielding also reduces fire or explosive hazards and minimizes any radio interference medium-voltage circuits might cause.
Electrostatic shielding of cables makes use of both nonmetallic and metallic materials. As shown in accompanying sketches of typical cable assemblies, semiconductive tapes or extruded coverings of semiconductive materials are combined with metal shielding to perform the shielding function. Metallic shielding may be done with:
1. A copper shielding tape wrapped over a semiconducting shielding of non-metallic tape that is applied over the conductor insulation (Fig. 310-9)
Fig. 310-9. A flat copper tape spiraled over the insulation is an electrostatic shield. (Sec. 310.6.)
2. A concentric wrapping of bare wires over a semiconducting, nonmetallic jacket over the conductor insulation (Fig. 310-10)
Fig. 310-10. Wires, instead of metal tape, are also used for electrostatic shielding (URD and UD type). (Sec. 310.6.)
3. Bare wires embedded in the semiconducting nonmetallic jacket that is applied over the insulation (Fig. 310-11)
4. A metal sheath over the conductor insulation, as with lead-jacketed cable
For many years, medium-voltage shielded power cables for indoor distribution circuits rated from 5 to 15 kV were of the type using copper tape shielding and an outer overall jacket. But in recent years, cables shielded by concentric-wrapped bare wires have also come into widespread use, particularly for underground outdoor systems up to 15 kV. These latter cables are the ones commonly used for underground residential distribution (called URD). Such a conductor is shown in Fig. 310-10.
In addition to use for URD (directly buried with the concentric-wire shield serving as the neutral or second conductor of the circuit), concentric-wire-shielded cables are also available for indoor power circuits, such as in conduit, with a nonmetallic outer jacket over the concentric wires. Such cable assemblies are commonly called drain-wire-shielded cable rather than concentric-neutral cable because the bare wires are used only as part of the electrostatic shielding and not also as a neutral. Smaller-gauge wires are used where they serve only for shielding and not as a neutral.
Fig. 310-11. Wires embedded in semiconducting jacket form another type of shielding. (Sec. 310.6.)
Figure 310-11 shows drain-wire-shielded medium-voltage cable with electro-static shielding by means of drain wires embedded in a semiconducting jacket over the conductor insulation. This type of drain-wire-shielded conductor is designed to be used for medium-voltage circuits in conduit or duct for commercial and industrial distribution as an alternative to tape-shielded cables. For the same conductor size, this type of embedded drain-wire-shielded cable has a smaller outside diameter and lighter weight than a conventional tape-shielded cable. For the drain-wire cable the assembly difference reduces installation labor, permits reduced bending radius for tight conditions and easier pulling in conduit, and affords faster terminations (with stress cones) and splices. An extremely important result of the smaller overall cross-sectional area (csa) of the drain-wire-shielded cable is the chance to use smaller conduits—with lower material and labor costs—when conduit is filled to 40 percent of its csa based on the actual cable csa, as covered by Note 5 to the tables in Chap. 9 of the NE Code.
Another consideration in conductor assemblies is that of strand shielding. As shown in Fig. 310-12, a semiconducting material is tape-wrapped or extruded onto the conductor strands and prevents voids between the insulation and the strands, thereby reducing possibilities of corona cutting on the inside of the insulation.
Refer to 300.40 on terminating and grounding shielded conductors.
Fig. 310-12. Strand shielding is part of the overall electrostatic shielding system on the conductor. (Sec. 310.6.)
310.7. Direct Burial Conductors. The second sentence in this section appears to require shielding only on direct burial cables rated above 2000 V, although rules in Sec. 300.50 say without reference to voltage that: “Direct burial cables shall comply with 310.7”; and also 300.50(A)(1) recognizes “nonshielded Type MC.” Put together, this correlates with terms of Exception No. 1 here that conditions unshielded cables in the 2 to 2.4 kV range (the former 5 kV allowance here gone the same way as the former 8 kV allowance mentioned above) “on being configured as multiconductor cable” (see Fig. 310-13), and having “an overall metallic sheath or armor.” Therefore, the default position of the NEC is that direct-burial cables over 2 kV must be shielded, with an exception for multiconductor cables up to 2.4 kV if the cable has an overall metallic sheath or armor. There is no NEC listing requirement for this cable, which is different from the aboveground requirement in 310.6. However, the section has generic language in its parent text requiring these direct-burial cables to be “identified for such use.” Review the discussion in this book at the definition of “identified” in Art. 100 for more information on this point.
Fig. 310-13. This application is covered by both Secs. 310.7 and 300.50. (Sec. 310.7.)
There is also an exception to retain the former allowance for unshielded 5-kV cabling for direct burial applications of series-connected airport runway lighting. Series lighting is still used in some locations by utilities to provide street lighting, although it is disappearing. It is widely used, with FAA sanction, for airport runways, in applications that are squarely premises wiring and not protected by the utility exemption.
The first sentence of this section applies to conductors rated for 600 V and lower systems as well as medium-voltage systems, and as noted, it requires conductors used for direct-burial applications to be identified for this purpose. The remainder of the Code rules on underground use of conductors rated up to 600 V are covered in the following paragraphs..
Figure 310-14 shows a clear violation of 310.7. In the photo, conductors marked RHW are run from the junction box below the magnetic contactor, directly buried in the ground. Although Type RHW is suitable for wet locations, it is not approved for direct burial. If, however, the conductors were of the type that is marked RHW-USE—that is, it is listed and recognized as both a single-conductor RHW and a single-conductor Type USE (underground service entrance) cable—then such conductors would satisfy this section.
Fig. 310-14. Bundle of conductors (arrow) are Type RHW individual building conductors that would be suitable for installation in conduit underground but are not marked “USE”; and their use here, run directly buried to outdoor lighting poles, constitutes a violation of the last sentence of Sec. 310.8. (Sec. 310.8.)
Type UF cable is acceptable for direct earth burial. Section 338.10(B)(4)(b) says USE cable used for branch circuits and feeders is to be installed in accordance with Part II of Art. 340 on Type UF cable, and 340.10(A) specifically says the cable is acceptable for direct burial, so that concludes the issue. In addition, the UL Electrical Construction Materials Directory (Green Book) notes that listed USE cable is recognized for burial directly in the earth (Fig. 310-15).
Fig. 310-15. Types USE and UF cables are designated by the letter U for underground use. (Sec. 310.8.)
Besides UF and USE, then, what other cables can be directly buried? 330.10(A)(5) does recognize MC cable for direct burial “when identified for such use.” 332.10(10) recognize MI cable for direct burial. Note that 332.10(6) permits MI in “fill” below grade.
For burial-depth requirements on directly buried cables, refer to 300.5 and Table 300.5. Cables approved for direct earth burial must be installed a minimum of 24 in. (600 mm) below grade, as given in Table 300.5, or at least 30 in. (750 mm) below grade for medium-voltage cables as covered in Table 300.50 with its exceptions.
Direct burial conductors should be trench-laid without crossovers; should be slightly “snaked” to allow for possible earth settlement, movement, or heaving due to frost action; and should have cushions and covers of sand or screened fill to protect conductors against sharp objects in trenches or backfill. Figure 310-16 shows some recommended details on installing direct burial cables. Moreover, when conductors are routed beneath roadways or railroads, they should be additionally protected by conduits. To guard against damage that might occur during future digging, conductors in soft fill should be covered by concrete slabs or treated planks.
Where prewired cable-in-conduit is being buried, it also should be slightly snaked, although it is unnecessary to provide sand beds or screen the backfill. Inasmuch as these complete conductor-raceway assemblies can be delivered on reels in specified factory-cut lengths, installation is simplified and expedited.
Fig. 310-16. This satisfies the intent of Sec. 300.5(F). (Sec. 310.7.)
310.8. Locations. Part (A) on dry locations simply says to use wires as specified in the Code. Part (B) on dry and damp locations lists insulation types that generally, but not always, have a “W” in their insulation designation. Part (C) covers wet locations. Any conductor used in a “wet location” (refer to the definition under “location” in Art. 100) must be one of the designated types—each of which has the letter W in its marking to indicate suitability to wet locations. Any conduit run underground is assumed to be subject to water infiltration and is, therefore, a wet location, requiring use of only the listed conductor types within the raceway. In addition, “moisture-impervious metal-sheathed” cables and conductors can be used; until the 1999 NEC this was a permission to use lead covered wire; the reason for making the change in terminology was never documented. Finally, any type listed for use in wet locations, even if not on the specific insulation list, can be used as well.
Part (D) covers locations exposed to direct sunlight, and requires that wiring directly exposed to the sun to have been evaluated for that exposure, or be protected by tape or sleeving that has been so evaluated. UL does those tests as part of its listing evaluations for Type SE cable so there isn’t a problem there. Incidentally, SE cable is an example of wiring that is listed sunlight resistant but the individual wires are not marked accordingly. If they were, then by Code usage of the cable would be “listed and marked”; instead it qualifies as “listed.” Either one is OK. However, on conduit services this whole requirement has been controversial, since almost every drip loop ever installed since before there even was an NEC (1897) has used whatever building wire was on hand, provided it was good for wet locations. There have been a number of proposals to exempt drip loops, which emerge from separately bushed holes in the weatherhead, to no avail. Listed coverings are beginning to find their way to market, as are conductors. One major manufacturer is now selling all of its RHH/RHW-2/USE-2 as sunlight resistant, plus its XHHW-2 and its THHN/THWN-2 in all sizes of 2 AWG and larger as having been listed and marked as sunlight resistant.
310.9. Corrosive Conditions. Figure 310-17 shows how conductors such as Type THHN-THWN are marked to indicate that they are gasoline- and oil-resistant, for use in gasoline stations and similar places.
Fig. 310-17. Typical marking indicates suitability of conductors for use under unusual environmental conditions. (Sec. 310.9.)
310.10. Temperature Limitation of Conductors. This rule says that no matter how your theoretical calculations seem to come out, no matter what number an ampacity table might seem to say is justified, if at the end of the day a wire will be consistently operating at a temperature above the ultimate temperature rating of its insulation, its operating conditions must change accordingly. The factors that influence this are set out in the detailed note following the rule. This rule is seldom cited by electrical inspectors. Instead, it is usually cited indirectly, such as through enforcement of 310.15(B)(2)(a) around a failure to allow for mutual conductor heating (thereby implicitly applying list item 4 from the note), or the failure to account for an elevated ambient temperature derating factor when using the pertinent ampacity table through enforcement of 240.4 (thereby implicitly applying list item 1 from the note), or a failure to account for heating in a neutral in response to triplen harmonic loading through enforcement of 310.15(B)(4)(c) (thereby implicitly applying list item 2 from the note), etc.
However, there are circumstances where this rule must be cited directly, because no other rule in the NEC can be cited. What follows is a real-world example, together with the underlying research that made it come together. This author was called on to inspect a single-family whole-house renovation that included a feeder from the basement to a panel in a second floor storage area. The feeder was protected with a 100 A circuit breaker, and was wired with 2/3 aluminum SER cable with a 4 AWG equipment grounding conductor. The feeder ran through an outside wall of the building, where it went up two stories before swinging over through the attic floor joists and down to the panel. The house had been completely gutted and entirely new insulation consisting of polyurethane foam had been sprayed into all outside wall and ceiling joist cavities, completely embedding this cable.
Now to begin with, since the feeder was not carrying the full load of this single-family house, the special ampacity table in 310.15(B)(6) does not apply, and therefore the normal values in Table 310.16 apply, which in this case would be 90 A, because the 75°C column must be used per 110.14(C)(1)(b). There were no continuous loads on this feeder. However, reducing the feeder protection to 90 A, although fixing the preceding violations, would still have completely failed to cure a blatant violation of 310.10. The reasons for this make good reading.
When wiring is embedded in thermal insulation, the ampacity is significantly degraded, and foamed-in-place polyurethane foam has about the best “R” value of any commercially available thermal insulation. The results tend to get progressively worse for larger cable sizes, because larger cables are generally installed to meet an expectation of higher current draw. The heat generated by a conductor is I2 R. Larger cables carrying larger currents reduce the heat by the first power due to larger radiating surface areas, and increase the heat by the square of the current increase. This means that large cables tend to be worse than smaller ones, because the latter increase in heat (+
I2) usually overwhelms the former decrease (- R).
In the 1987 NEC cycle, this was confirmed by actual NEMA testing, with dramatic results. The test used 2 AWG aluminum Type SE cable of the “SEU” style (two insulated conductors and the grounded conductor configured as a spirally-wrapped conductor around the ungrounded conductors). This cable is used routinely for 100 A residential services. It was run through thermal insulation under controlled test conditions. Specifically, the cable was run embedded in cellulose insulation with 7 in. above it. Thermocouples were placed on the cable, and the various loads under test were maintained for long periods. This cable has a Table 310.16 ampacity (terminations not considered for this purpose) of 100 A. When the cable was loaded to 100 A, the cable jacket was “completely charred” as well as adjacent “charred wood members,” all while the cable was operating within its table ampacity limitations. In fact, the testing showed (65 A caused 96°C operation) the cable exceeded its rated operating temperature at any time the continuous current exceeded about two-thirds of the table ampacity.
These results strongly demonstrate the fact that the ampacity of a conductor is not necessarily what a table may predict. The ampacity of a conductor is its ability to carry current continuously under the conditions of use. This is determined by thermodynamics, not code tables. The ampacity of 2 AWG aluminum XHHW configured as 2/3 Al SEU cable and embedded in cellulose insulation is approximately 60 A. Actually, the ampacity corresponding to Table 310.16 would be even less than that, because Table 310.16 is structured around three current-carrying conductors and this testing only used two of the three conductors in the test circuit.
This testing focused on Type SE cable, but a test run using 6 AWG copper in metal conduit embedded in foam insulation overheated those conductors as well, although not as badly. Typical branch circuits do OK by starting in the usually mandated 60°C ampacity column. There is no ampacity table for thermal insulation, and it would be impossible to create one, because variations in R factors significantly change end results. However, it is fair to say, for example, that the ampacity of 4/0 Al SE cable (table ampacity of 205 A) embedded in thermal insulation is far less than even the 60°C table result of 150 A would predict, perhaps as low as 100 A. NEC 310.10, now focusing on item 3 of its application note, is the only NEC rule that covers this point, and it is referenced in notes to a number of cable articles for this reason. To the extent practicable, good practice is to route larger circuits in such a way as to avoid contact with thermal insulation.
And so the citation was 310.10, cured with a 60 A circuit breaker, whose terminals were fortunately large enough to accept the 2 AWG feeder conductors, that ran to a panel where, fortunately again, the load per Art. 220 was probably not over 35 A.
310.12. Conductor Identification. For part (A), refer to the discussion given for 200.6 and 200.7. For part (B), refer to 250.119(A) and (B). 310.12(A) and 200.6(E) recognize the use in multiconductor cables of a grounded conductor that is not white throughout its entire length, provided that only qualified persons will service the installation. The rule requires that such grounded conductors be identified by white marking at their termination at the time of installation.
Similarly, a grounding conductor in a multiconductor cable may be identified at each end and at every point where the conductor is accessible by stripping the insulation from the entire exposed length or by coloring the exposed insulation green or by marking with green tape or green adhesive labels.
Part (C) for ungrounded conductors now simply defers to 210.5(C) for branch circuit and 215.12 for feeders on color (or other method of) coding. The method used cannot render the required conductor markings in 310.11(B)(1) unreadable.
310.13. Conductor Constructions and Applications. Table 310.13(A) presents application and construction data on the wide range of 600-V insulated, individual conductors recognized by the NE Code, with the appropriate letter designation used to identify each type of insulated conductor. Figure 310-18 shows a typical detail on application, as covered for Type THW conductor in NEC Table 310.13(A). Type THW wire has a special application provision for electric-discharge lighting, which makes THW an acceptable answer for installers needing a 90°C conductor for wiring end-to-end fixtures in compliance with Sec. 410.65.
Fig. 310-18. THW wire has the 90°C rating required of conductors within 3 in. (76 mm) of a ballast. (Secs. 410.31 and 310.13.)
Important data that should be noted in Table 310.13(A) are as follows:
1. The designation for “thousand circular mils” is “kcmil,” which has been substituted for the long-time designation “MCM” in this table and throughout the NEC.
2. Type MI (mineral insulated) cable may have either a copper or an alloy steel sheath.
3. Type RHW-2 is a conductor insulation that is moisture- and heat-resistant rubber with a 90°C rating, for use in dry and wet locations.
4. Type XHHW-2 is a moisture- and heat-resistant cross-linked synthetic polymer with a 90°C rating, for use in dry and wet locations.
5. The suffix “LS” designates a conductor insulation to be “low smoke” producing and flame retardant. For example, Type THHN/LS is a THHN conductor with a limited smoke-producing characteristic.
6. Type THHW is a moisture- and heat-resistant insulation, rated at 75°C for wet locations and 90°C for dry locations. This is similar to THWN and THHN without the outer nylon covering but with thicker insulation.
7. All insulations using asbestos—A, AA, AI, AIA, AVA, etc.—have been deleted from Table 310.13(A) because they are no longer made.
8. A new provision (2008) formally recognizes the long-standing manufacturing practice of segmenting the equipment grounding conductor in large MC and TC cable assemblies, and others, to make the cable more workable and easier to produce. The total of the area of the resulting segments must not be less than the size required by 250.122.
Conductors for medium-voltage circuits (over 2000 V) must satisfy Tables 310.13(B) through (E) as applicable, and the ampacity values given in Tables 310.67 through 310.86 subject to the conditions in 310.60.
Conductors intended for 600-V (and up to 2000-V) general wiring under the requirements of the National Electrical Code are required to be one of the recognized types listed in Code Table 310.13(A) and not smaller than No. 14 AWG. The National Electrical Code does not contain detailed requirements for insulated conductors since these are covered in separate standards such as those of Underwriters Laboratories Inc.
“Dry locations” in this case would mean for general use in dry locations.
Table 310.13(A) permits maximum operating temperatures of 90°C (194°F) in dry and damp (but not “wet”) locations for Types FEP, FEPB, RHH, XHHW, and THHN wire; but the load-current ratings for 14, 12, and 10 AWG copper conductors and 12 and 10 AWG aluminum conductors are limited to those permitted by the maximum overcurrent protection ratings given in 240.4(D). One reason is the inability of 15-, 20-, and 30-A CBs to protect these sized conductors against damage under short-circuit conditions. The other reason is that the wiring devices that are commonly connected by these sizes of conductors are not suitable for conditions encountered at higher current loadings.
Terminals of 15- and 20-A receptacles not marked CO/ALR are for use with copper and copper-clad aluminum conductors only. Terminals marked CO/ALR are for use with aluminum, copper, and copper-clad aluminum conductors. Screwless pressure terminal connectors of the conductor push-in type are for use only with copper and copper-clad aluminum conductors.
Terminals of receptacles rated 30 A and above not marked AL-CU are for use with copper conductors only. Terminals of receptacles rated 30 A and above marked AL-CU are for use with aluminum, copper, and copper-clad aluminum conductors.
The conductor material known as copper-clad aluminum is made from a metallurgical materials system by using a core of aluminum with a bonded outer skin of copper. There is 10 percent copper by volume (the outer skin) and 26.8 percent by weight. Terminations for copper-clad aluminum conductors should be marked AL-CU, except where listings by Underwriters Laboratories indicate otherwise.
Wire connectors have limitations with respect to their application on stranded wire. Unless provided with specific markings (such as “SOL” or “STR”), wire connectors for 30 through 10 AWG sizes can be used for both stranded and solid conductors, and wire connectors rated for 8 AWG and larger wires are for stranded conductors only. Connectors used for aluminum stranded conductors can be used for Class B concentric, compressed, and unidirectional lay compact stranding. Connectors used for stranded copper terminations are suitable for Class B concentric and compressed conductors, and also Class C concentric conductors. Connectors used for compact-stranded copper are marked for that purpose either on the connector or on the smallest unit carton in which the connectors are sold. Connectors suitable for more finely stranded conductors, such as Class H or Class M stranding, are marked with the additional stranding class designation and also the number of strands. Most listed connectors for fine-stranded conductors are in the form of crimp-on lugs that force the strand ends together into a homogeneous mass of metal.
310.15. Ampacities. 310.15 states that ampacities of conductors may be determined by either of two methods. The first method is described in part (B) and is the old, tested, and familiar method of the NEC, based on Tables 310.16 through 310.21. The second permitted method is covered in part (C) of the NEC and is the complex, confusing, incomplete, and problematic procedure that was presented in the 1987 NEC as the basic method, based on an elaborate formula to be applied under very sophisticated engineering supervision, as stated there.
The NEC ampacity determination procedure using the formula is permitted as an optional alternative “under engineering supervision.” All of the ampacity tables based on the formula are in App. B in the back of the Code book, where information on the formula method and its related ampacity tables is introduced with the sentence, “This appendix is not part of the requirements of this Code, but is included for information purposes only.” Thus, the 1987 NEC ampacity method is given as a nonmandatory, optional alternative to the old standby method. Included in Annex B, however, is the old Note 8 ampacity reduction factors (Table B-310.11) for multiple conductors in a raceway.
In part (B), a fine-print note (FPN) points out that “Tables 310.16 through 310.19 are application tables that are for use in determining conductor size on loads calculated in accordance with Article 220.” Inasmuch as the NEC itself requires that Art. 220 be used at all times in calculating loads, the ampacity-determination method of part (B) is completely adequate for all conductor sizing in accordance with all Code rules.
Correctly integrating load calculation results with ampacity table values, ambient temperature adjustment factors, mutual conductor heating derating factors, termination temperature limitations, etc., is quite possibly the most challenging task confronting electrical personnel today, and also one of the most necessary. After all, if we do nothing else in this trade, we select and install wires. Since there is now a comprehensive example that focuses on these issues at the end of the Code in Annex D [Example D3(a)], the place to integrate these rules is there. Refer to the comprehensive discussion of these issues in Chap. 9 of this book. In addition, there is an analysis with step-by-step calculations on Table B-310.11 so you can see how to bring the old Note 8 back to life.
(B)(6) This rule and table have been in the NE Code for a long time and permit use of certain conductors at ampacity values higher than those shown for the conductors in Table 310.16. For instance, a 2/0 AWG THW copper conductor may be used protected at 200 A instead of at 175 A, as shown in Table 310.16. This permission has been given by the NE Code in recognition of the reality that residential service conductors are supplying loads of great diversity and of short operating periods or cycles, so that the conductors almost never see full demand load approaching their ampacity and certainly not for continuous operation (3 h or more).
The higher allowable OC device protective ratings for 120/240-V, 3-wire (not 2-phase and neutral from a 208Y/120-V system) single-phase dwelling services may also be utilized for feeder conductors, and Type USE is one of the types of conductors, given the higher ampacities. The ampere rating of services that are subject to the higher ampacity values was increased from 200 to 400 A.
Those higher ratings of protective devices permitted by part (B)(6) are applicable only for service conductors or feeder conductors used in “dwelling units,” thereby limiting the application to one-family houses and individual apartments in apartment houses, condominiums, and the like, because only such units conform to the Code definition of “dwelling unit,” as given in Art. 100. The wording excludes use of part (B)(6) in relation to service-entrance conductors to a whole building, such as a multifamily dwelling, as shown in Fig. 310-19. And in that drawing, prior to the 1990 Code, the wording did prohibit use of the higher ampacities of conductors if each of the SE runs to the individual apartments were a feeder instead of service-entrance conductors, as each run would be if a disconnect and protective device were used at each meter location. But reason dictates that the higher ampacities should be allowed in either case, even though each feeder does not carry “the total current supplied by that service.” The phrase “total current supplied by that service” was removed in the 1990 NEC, which has the effect of permitting the use of the increased ampacities for feeders to individual “dwelling units” in a multifamily dwelling. And, effective with the 2008 NEC, there is now express language forbidding the use of these allowances for the service-entrance conductors to the whole building.
Prior to the 1990 NEC, in Note 3, [now subpart (6) of 310.15(B)], it was clearly indicated that the higher allowable ampacities for residential occupancies using 3-wire, single-phase services also may be applied to 3-wire, single-phase feeders in those cases where the feeder conductors from the service equipment to a subpanel or other distribution point carry the total current supplied by the service conductors. Inclusion of the phrase “that supply the total load to a dwelling unit” in the 1993 Code would not recognize use of the increased ampacities given in Note 3 for a feeder as shown in the one-family house drawing of Fig. 310-19, if, say, swimming pool circuits or other loads are supplied from the service but are not carried by the feeder. Another revision in this part of the 2008 NEC expressly prohibits application of 310.15(B)(6) to a feeder that does not see the entire dwelling load, and thereby benefit from the large diversity that comes with that load.
The last sentence of 310.15(B)(6) permits the neutral conductor of these 3-wire services and 3-wire feeders to be smaller than the hot conductors because the neutral carries only the unbalanced current of the hot legs and is not at all involved with 2-wire, 240-V loads. However, it must not be sized smaller than 215.2, 220.61, and 230.42 requirements.
Fig. 310-19. Higher ratings of protective devices may be used for service conductors to “dwelling units.”
Part (B)(3) 310.15(B)(3) provides that if an uninsulated conductor is used with insulated conductors in a raceway or cable, its size shall be the size that would be required based on it being a conductor having the same insulation as the lowest insulation temperature rating of the adjacent insulated conductors. The point is to assure that its surface temperature will not exceed the temperature rating of any adjacent conductor insulation (Fig. 310-20).
example Two 6-AWG Type THW conductors and one bare 6-AWG conductor in a race-way or cable. The ampacity of the bare conductor would be 65 A.
If the insulated conductor were Type TW, the ampacity of the bare conductor would be 55 A.
Part (B)(2) Where more than three current-carrying conductors are used in a raceway or cable, their current-carrying capacities must be reduced to compensate for the proximity heating effect and reduced heat dissipation due to reduced ventilation of the individual conductors that are bunched or form an enclosed group of closely placed conductors. Where the number of conductors in a race-way or cable exceeds three, the ampacity of each conductor shall be reduced as indicated in the table of part (B)(2).
Fig. 310-20. How to figure ampacity of a bare conductor, where permitted. (Sec. 310.15.)
If, for instance, four 8 AWG THW copper conductors are used in a conduit, the ampacity of each 8 AWG is reduced from the 50-A value shown in the table to 80 percent of that value. In such a case, each 8 AWG then has a new reduced ampacity of 0.8 × 50 A, or 40 A. And, from 240.4, “Conductors shall be protected in accordance with their ampacities.” Thus, 40-A-rated fuses or CB poles would be required for overcurrent protection as the general rule.
The application of those part (B)(2) conductors and their protection rating is based on the general concept behind the NE Code tables of maximum allowable current-carrying capacities (called “ampacities”). The NE Code tables of ampacities of insulated conductors installed in a raceway or cable have always set the maximum continuous current that a given size of conductor can carry continuously (for 3 h or longer) without exceeding the temperature limitation of the insulation on the conductor, that is, the current above which the insulation would be damaged. But, because the overcurrent devices were tested with conductors sized at 125 percent of the continuous current plus the noncontinuous, the conductor’s prederated ampacity must be increased where supplying a continuous load [see 210.19(A)].
This concept has always been verified in the FPN to 240.1, where the wording has been virtually identical for over 30 years and says, “Overcurrent protection for conductors and equipment is provided to open the circuit if the current reaches a value that will cause an excessive or dangerous temperature in conductors or conductor insulation.” To correspond with that objective, 240.4 says, “Conductors, other than flexible cords and fixture wires, shall be protected against overcurrent in accordance with their ampacities as specified in 310.15.”
Table 310.16, for instance, gives ampacities under the conditions that the raceway or cable containing the conductors is operating in an ambient not over 30°C (86°F) and that there are not more than three current-carrying conductors in the raceway or cable. Under those conditions, the ampacities shown correspond to the thermal limit of the particular insulations. But if either of the two conditions is exceeded, ampacities have to be reduced to keep heat from exceeding the temperature limits of the insulation:
1. If ambient is above 30°C, the ampacity must be reduced in accordance with the correction factors given at the bottom of Table 310.16.
2. If more than three current-carrying conductors are used in a single cable or raceway, the conductors tend to be bundled in such a way that their heat-dissipating capability is reduced and excessive heating would occur at the ampacities shown in the table. As a result, part (B)(2) requires reduction of ampacity, and conductors must be protected at the reduced ampacity.
(Note: It should be clearly understood that any reduced ampacity—required for higher ambient and/or conductor bundling—has the same meaning as the value shown in a table: Each represents a current value above which excessive heating would occur under the particular conditions. And if there are two conditions that reduce heat dissipation, then more reduction of current is required than for one condition of reduced dissipation.)
Part (B)(2)(a) says, “Where the number of conductors in a raceway or cable exceeds three . . . the ampacities shall be reduced as shown in the following table.” Table 310.15(B)(2)(a) has a heading on the right to require that any ampacity derating for elevated ambient temperature must be made in addition to the one for number of conductors. If, for instance, four No. 8 THHN current-carrying copper conductors are used in a conduit, the ampacity of each No. 8 is reduced from the 55-A value shown in Table 310.16 to 80 percent of that value. Each No. 8 then has a new (reduced) ampacity of 0.8 × 55 A, or 44 A. Then, if a derating factor must be applied because the conductors are in a conduit where the ambient temperature is, say, 40°C instead of 30°C, the factor of 0.91 (36 to 40°C) from the bottom of Table 310.16 must be applied to the 44-A current value to determine the final value of ampacity for the conductors (44 × 0.91 = 40 A). Moreover, 240.4 of the NEC states, “Conductors, other than flexible cords and fixture wires, shall be protected against overcurrent in accordance with their ampacities as specified in 310.15.” Thus, fuses or CB poles rated at 40 A would be required. The ampacity of the conductors is changed and the conductors must be protected in accordance with the derated ampacity value and not in accordance with the tabulated value.
Because conductor ampacity is reduced when more than three conductors are used in a conduit, the overcurrent protection for each phase leg of a parallel makeup in a single conduit would generally have to be rated at not more than the sum of the derated ampacities of the number of conductors used per phase leg. That would satisfy 240.4, which requires conductors to be protected at their ampacities. Because ampacity is reduced in accordance with the percentage factors given in part (B)(2) for more than three conductors in a single conduit, that derating dictates the use of multiple conduits for parallel-makeup circuits to avoid the penalty of loss of ampacity.
Figure 310-21 shows examples of circuit makeups based on the unsafe concept of load limitation instead of ampacity derating, as applied to overcurrent rating and conductor ampacity—which is a Code violation, because the conductors are not protected in accordance with their ampacities.
Fig. 310-21. Parallel-conductor makeup must not be used in single conduits without ampacity reduction, even if load current is limited to the conductor ampacity as shown here.
Figure 310-22 shows a condition of bunched or bundled (now clarified as “installed without maintaining spacing for a continuous length”) Type NM cables where they come together at a panelboard location. The rule of 310.15(B)(2) requires conductors in bundled cables to have their load currents reduced from the ampacity values shown in Table 310.16. The ampacity derating is required for conductor stacks or bundles that are longer than 2 ft (24 in. or 600 mm). For shorter bundles, derating is not required in general; however, note that Type NM cable has an effective length of not much more than zero if multiple cables run through draft- or fire-stopped holes in framing members. Exception No. 3 excludes the need for derating groups of four or more conductors installed in nipples not over 24-in. (600 mm) long.
Exception No. 4 to 310.15(B)(2) says that underground conductors that are brought up aboveground in a protective raceway [300.5(D)] do not require derating if not more than four conductors are used and if the protective conduit has a length not over 10 ft (3.0 m) “above grade.” The total length of race-way may exceed 10 ft (3.0 m). The phrase above grade clearly limits the length of protective conduit that may contain conductors without derating in accordance with Table 310.15(B)(2)(a). The 10-ft (3.0-m) length covers the length of 8 ft (2.5 m) above grade but not the 1½ ft (450 mm) into the earth given by 300.5(D) on conductors emerging from underground.
Fig. 310-22. If a large number of multi-conductor cables are bundled together in a stud space, capacity derating in accordance with part (B)(2) would be required. If the individual cables are spaced apart and stapled, then the conductors in the cables may be loaded up to their rated ampacity values from Table 310.16. (Sec. 310.15.)
310.15(B)(2) does not apply to conductors in wireways and auxiliary gutters, as covered in 366.23 and 376.22. Wireways or auxiliary gutters may contain up to 30 conductors at any cross section [excluding signal circuits and control conductors used for starting duty only between a motor and its starter in auxiliary gutters, 366.22(A)]. The total cross-sectional area of the group of conductors must not be greater than 20 percent of the interior cross-sectional area of the wireway or gutter. And load-limiting factors for more than three conductors do not apply to wireway the way they do to wires in conduit. However, if the derating factors from Table 310.15(B)(2)(a) are used, there is no limit to the number of wires permitted in a wireway or an auxiliary gutter. However, the sum of the cross-sectional areas of all contained conductors at any cross section of the wire-way must not exceed 20 percent of the cross-sectional area of the wireway or auxiliary gutter. More than 30 conductors may be used under those conditions.
Part (B)(2)(b) says that spacing between “conduits, tubing, or raceways shall be maintained.” The most significant thing about this provision is not what it says, but what it very nearly said, and was stopped from saying at the last minute.
When the NEC panel began to seriously question the adequacy of derating factors for mutual conductor heating in the 1987 Code edition, it didn’t stop with the new values (which are still in the NEC) in what is now (a). It also was going to add a table to what is now (b). This table would have mandated an additional set of derating factors for conduit and tubing run on trapeze hangers where the spacing was less than a full raceway diameter apart! The table was multidimensional in that the additional penalties would have applied not just side-to-side, but also vertically in the event more than one trapeze was stacked up the same set of threaded rods. For example, three trapeze hangers with six raceway runs each, and with less than a raceway diameter separation in both dimensions, would have produced an additional derating factor of 0.74. Even two conduits on a trapeze hanger, if the conduits were separated by less than a full raceway diameter, would have generated an additional factor of 0.94. There was also language that probably would have been applied as a complete prohibition against any separation less than a quarter of a raceway diameter.
UL, as it has so many times, saved the industry by running tests in the comment period showing that with any type of conventional conduit clamps on adjacent conduits in the trapeze hanger, there was enough separation and air flow so the mutual heating was negligible. Anyone with an ounce of common sense and field experience already knew that. Since no one is throwing bushel baskets of conduits on trapeze hangers without securing them with some kind of hardware, we can pretty much ignore (b). No loss experience was ever documented to support the inclusion of either the note or the derating factors in the original proposal. However, be advised that it is a code violation to throw bushel baskets of conduits on trapeze hangers without securing them individually with some kind of hardware that provides a modicum of spacing.
Part (B)(2)(c) is new in the 2008 NEC, and is based on substantial research surrounding the effects of heat re-radiating from rooftops on to “conduits” exposed to direct sun. Note that this phrasing should certainly be expanded to “tubing” to incorporate EMT and quite possibly large cable assemblies as well, such as industrial, feeder-sized applications of Type MC cables that often run across rooftops on similar supports. The new provision adds a table of ambient temperature correction factor modifications based on the wiring distance from the top of the roof surface to the bottom of the conduit. The ambient temperature to be taken through the ampacity correction factors at the bottom of Table 310.16 becomes the sum of the number in the table based on conduit position, and the normal ambient temperature for the geographic location. There is a note that suggests using the ASHRAE Handbook to look up ambient temperature data.
The ASHRAE publication is based on 30 years of data, and looks at three different percentiles for the 3-month period from June through August. The 0.4 percent number is the temperature that will be exceeded for 0.4 percent of the total time in a 1-month period. Since a month is 720 h, this temperature will be exceeded for 3 h a month. The tables also calculate 1.0 and 2.0 percent numbers the same way. The number usually employed for these purposes is the 2.0 percent number, corresponding to 14 h a month. In the example that follows, we will use Hartford, CT, for which the 2.0 percent number is 36°C (90°F). The Copper Development Association has an extensive extract of this data from many U.S. cities available online.
example An IMC conduit will supply a rooftop air-conditioning unit, routed 4.5 m (15 ft) across the roof running on 2 × 4 blocks mounted edgewise and sitting on nominal 1 in. pressure-treated boards that will be glued to the roof. The unit has a nameplate FLA of 120 A and it will be connected to a 480Y/277-V system using THHN/THWN-2 copper wire, with the conduit as the equipment grounding conductor. What size wire and conduit is required?
answer: The wire size must be initially selected on the basis of the 75°C column in Table 310.16 based on 110.14(C)(1)(b), but using 125 percent of the FLA per 440.32.
120 A × 1.25 = 150 A. Note that this is to be used as the operating current for code purposes.
The wire cannot be smaller than 1/0 AWG in order to not overheat the terminations. However, now we need to look at conductor ampacity. The ampacity of this wire is 170 A from the table before adjustments. The 2 × 4 block will be 90 mm (3½ in.) on edge, and the flat board adds another 19 mm (¾ in.) for a total of 109 mm (4¼ in.) above the roof. From Table 310.15(B)(2)(c) the extra temperature to be added is 17°C (30°F) for a total design temperature of 48°C (120°F). According to the correction factors at the bottom of Table 310.16, this results in a correction factor of 0.82. Since the table ampacity is 170 A, the ampacity on the roof will be 170 A × 0.82 = 139 A. The wire is too small, but would have been allowable under prior codes. Without the rooftop adjustment the ampacity correction factor is 0.96, for 163 A. The 17°C (30°F) adder lowered the resulting ampacity by 24 A.
The next size up, 2/0 AWG (table ampacity = 195 A), will work well: 195 A × 0.82 = 160 A.
Part (B)(4) In the determination of conduit size, neutral conductors must be included in the total number of conductors because they occupy space as well as phase conductors. A completely separate consideration, however, is the relation of neutral conductors to the number of conductors, which determines whether ampacity derating must be applied to conductors in a conduit, as follows.
Neutral conductors that carry only unbalanced current from phase conductors (as in the case of normally balanced 3-wire, single-phase or 4-wire, 3-phase circuits supplying resistive loads) are not counted in determining ampacity derating of conductors on the basis of the number in a conduit, as described. A neutral conductor used with two phase legs of a 4-wire, 3-phase system to make up a 3-wire feeder is not a true neutral in the sense of carrying only current unbalance. Such a neutral carries the same current as the other two conductors under balanced load conditions and must be counted as a phase conductor when more than three conductors in conduit are derated.
The technical basis for the full-current loading on the 3-wire feeder neutral from a wye system is based on phasor analysis that shows that if you have a circuit consisting of two phase conductors and a neutral from a wye distribution with 100 A of line-to-neutral current on each phase conductor, perfectly balanced, the neutral will carry 100 A as well. Note that the identical load profile connected to a single-phase distribution would result in a zero ampere loading on the neutral. Further, as the load on one of the phase conductors in this example decreases, the load on the neutral does decline slightly, reaching a minimum of about 87 A (½ 
3 × highest load) and then rising again to 100 A when the load on the more lightly loaded phasor reaches zero and we are left with 100 A in what has become effectively a two-wire circuit. Therefore, these neutrals are effectively fully loaded at all times and must be counted accordingly. By the way, the minimum loading in these cases occurs when one phasor carries 100 A and the other is carrying one-half that amount or 50 A.
Because the neutral of a 3-phase, 4-wire wye branch circuit or feeder to a load of fluorescent, metal-halide, mercury, or sodium lamp lighting or to electronic data processing equipment—the so-called information technology equipment—or any other nonlinear load will carry harmonic current even under balanced loading on the phases (refer to 220.61), such a neutral is not a true noncurrent-carrying conductor and must be counted as a phase wire when the number of conductors to arrive at an ampacity derating factor is determined for more than three conductors in a conduit. As a result, all the conductors of a 3-phase, 4-wire branch circuit or feeder to a fluorescent load would have an ampacity of only 80 percent of their nominal ampacity from Table 310.16 or other ampacity table. Because the 80 percent is a derating of ampacity, the conductors must be protected at the derated ampacity value.
Figure 310-23 shows four basic conditions of neutral loading and the need for counting the neutral conductor in loading a circuit to fluorescent or mercury ballasts, as follows:
Case 1—With balanced loads of equal power factor, there is no neutral current, and consequently no heating contributed by the neutral conductor. For purposes of heat derating according to the Code, this circuit produces the heating effect of only three conductors.
Case 2—With two phases loaded and the third unloaded, the neutral carries the same as the phases, but there is still the heating effect of only three conductors.
Case 3—With two phases fully loaded and the third phase partially loaded, the neutral carries the difference in current between the full phase value and the partial phase value, so that again there is the heating effect of only three full-load phases.
Case 4—With a balanced load of fluorescent ballasts, third-harmonic current generation causes a neutral current approximating phase current, and there will be the heating effect of four conductors. Such a neutral conductor must be counted with the phase conductors when the load-current limitation due to conduit occupancy is determined, as required in part (C) of part (B)(4).
Although part (B)(4) exempts only neutral conductors from those conductors that must be counted in determining load-limiting factors for more than three conductors in a raceway or cable [per part (B)(2)], similar exemption should be allowed for one of the “travelers” in a 3-way (or 3- and 4-way) switch circuit. As shown in Fig. 310-24, only one of the two conductors is a current-carrying conductor at any one time; therefore, the other should not be counted for load-limitation purposes where such switch legs are run in conduit or EMT along with other circuit conductors. No change in code language is needed to accomplish this; when one of two wires in a conduit necessarily carries zero amperes at all times, even if the wire at zero switches its identity from time to time, the number of current-carrying conductors countable in that group of two is: exactly one.
Fig. 310-23. All neutrals count for conduit fill, but only “true neutrals” do not count in determining ampacity derating for number of conductors in a raceway or cable [Part (B)(4)(b)]. (Sec. 310.15.)
Fig. 310-24. The 3-wire run in conduit between 3-way switches contains only two current-carrying conductors. (Sec. 310.15.)
Part (B)(5) This note makes it clear that an equipment grounding conductor or bonding conductor, which under normal conditions is carrying no current, does not have to be counted in determining ampacity derating of conductors when more than three conductors are used in a raceway or cable. As a result, equipment grounding and bonding conductors do not have to be factored into the calculation of required ampacity derating specified in part (B)(2).
Ampacity of conductors over 600 V, nominal Since the 1975 edition, the NEC has added a vast amount of information and data for conductors rated over 600 V, up to 35 kV. For instance, Tables 310.67 through 310.86 give maximum continuous ampacities for copper and aluminum solid dielectric insulated conductors rated from 2001 to 35,000 V. Note that there are two ampacity columns in the tables, one for 90°C and one for 105°C. This sets up a comparable situation to the one at 600 V and below where there are termination rules and ampacity rules. In this case, however, the terminations take on the 90°C column per 110.40 regardless of conductor choice, but the 105°C numbers can be used as the starting point for ampacity calculations based on environmental factors.
312.1. Scope. Cabinets and cutout boxes, according to the definitions in Art. 100, must have doors (cutout boxes can also have telescoping covers), and are thus distinguished from large boxes with covers consisting of plates attached with screws or bolts. Article 312 applies to all boxes used to enclose operating apparatus—that is, apparatus having moving parts or requiring inspection or attention, such as panelboards, cutouts, switches, circuit breakers, control apparatus, and meter socket enclosures.
312.2. Damp and Wet Locations. This section addresses “damp and wet locations.” The Code requires all equipment covered by this article to be “placed or equipped” so as to ensure water does not enter or collect within the enclosure. Additionally, a minimum clearance from the mounting surface of ¼ in. must be maintained to prevent corrosion of the enclosure. And, fittings at raceway or cable entries in metallic enclosures must be listed for use in “wet locations” where such entry is above the level of any uninsulated live parts. This qualification reflects the fact that any energized part, whether insulated or not, is a live part per the Art. 100 definition. The clearance to the mounting surface is not required for non-metallic enclosures on nonabsorbent surfaces, as noted by the Exception. The rule about wet location fittings being required for entries above a certain level answers a long-standing question as to whether a conventional SE cable connector suitable for dry locations can be used with SE cable exiting from the bottom of an outdoor meter socket. The answer is that it can be used, but the fitting for the cable entry in the top of the socket does have to be listed for wet locations.
312.3. Position in Wall. Figure 312-1 shows how the¼-in. (6.35-mm) setback relates to cabinets installed in noncombustible walls.
312.4. Repairing Noncombustible Surfaces. The gap between the enclosure that uses a flush-mounted cover and adjacent finished surfaces—where the surface is plaster, drywall, or plasterboard—shall not be greater than 3 mm (
in.). This makes the rule identical to the comparable rule in 314.21. Both 312.3 and 312.4, and their counterparts in Art. 314, address issues related to when ordinary building construction will be allowed to complete an electrical enclosure. Review the commentary following 314.21 for more information regarding the intent behind these requirements.
Fig. 312-1. In masonry wall, cabinet does not have to be flush with wall surface—as it does in wood wall. (Sec. 312.3.)
312.5. Cabinets, Cutout Boxes, and Meter Socket Enclosures. Part (C) makes clear that all cables used with cabinets or cutout boxes must be attached to the enclosure. NM cable, for instance, does not have to be connected by clamp or connector device to a single-gang nonmetallic outlet box as in 314.17(C), but must always be connected to KOs in panelboard enclosures and other cabinets (Fig. 312-2). If installations require concealed wiring, spare conduits should be stubbed to accessible areas, such as above the lift-out ceiling-panel construction in common use today. In such areas this section includes an allowance for running multiple nonmetallic cable assemblies through a vertical riser. This facilitates the use of Type NM cables for light commercial construction; a surface panel can go on the wall at a convenient point with a single large riser conduit stub poked through the suspended ceiling to receive the cables. There are numerous limitations to be observed, but the neat and workmanlike result often makes this approach worthwhile. Specifically, the installation must meet the following conditions:
The cables must have entirely nonmetallic sheaths.
The raceway must enter the top of the cabinet.
The raceway must be nonflexible, such as EMT.
The raceway must be no shorter than 450 mm (18 in.), nor longer than 3.0 m (10 ft).
Every cable entering the raceway end must be secured not over 300 mm (12 in.) from the entry point to the raceway.
The raceway must extend directly above the cabinet.
The raceway must not penetrate a structural ceiling.
The raceway must be equipped with fittings on both ends to protect the cables from abrasion.
The outer raceway end must be sealed or plugged by an approved means to prevent debris from falling into the cabinet. Duct seal is often used for this purpose.
The cables that run to the enclosure must have their sheaths intact through the entire raceway section.
The raceway must be supported in accordance with the rules that apply to the wiring method employed.
The cable-fill limitations must be observed. This means the cables cannot occupy more than 40 percent of the raceway internal cross section, and for elliptical cables, the cable cross section is taken as the area of a circle whose diameter equals the major axis of the ellipse.
Fig. 312-2. All cables must be secured to all cabinets or cutout boxes. (Sec. 312.5.)
The mutual conductor heating rules, as covered in 310.15(B)(2)(a), must be applied. This is perhaps the most problematic constraint; however, if the raceway riser does not exceed 600 mm (24 in.) in length, those rules can be ignored because the raceway now qualifies as a nipple.
312.6. Deflection of Conductors. Parts (A) and (B) cover a basic Code rule that is referenced in a number of Code articles to ensure safety and effective conductor application by providing enough space to bend conductors within enclosures.
A basic concept of evaluating adequate space for bending conductors at terminals of equipment installed in cabinets is presented in this section. The matter of bending space for conductors at terminals is divided into two different configurations, as follows:
1. The conductor does not enter (or leave) the enclosure through the wall opposite its terminals. This would be any case where the conductor passes through a wall of the enclosure at right angles to the wall opposite the terminal lugs to which the conductor is connected or at the opposite end of the enclosure. In all such cases, the bend at the terminals is a single-angle bend (90° bend), and the conductor then passes out of the bending space. It is also called an L bend, as shown at the top left of Fig. 312-3. For bends of that type, the distance from the terminal lugs to the wall opposite the lugs must conform to Table 312.6(A), which requires lesser distances than those of Table 312.6(B) because single bends are more easily made in conductors, and they are easier to land on a terminal.
2. The conductor enters (or leaves) the enclosure through the wall opposite its terminals. This is a more difficult condition because the conductor must make an offset or double bend to go from the terminal and then align with the raceway or cable entrance. This is also called an S or a Z bend because of its configuration, as shown at the top right of Fig. 312-3. For such bends, Table 312.6(B) specifies a greater distance from the end of the lug to the opposite wall to accommodate the two 45° bends, which are made difficult by the short lateral space between lugs and the stiffness of conductors (especially with the plastic insulations in cold weather).
Table 312.6(B) provides increased bending space to accommodate use of factory-installed connectors that are not of the lay-in or removable type and to allow use of field-installed terminals that are not designated by the manufacturer as part of the equipment marking. Exception No. 1 to part (B) is shown in the bottom drawing of Fig. 312-3. This setup is often used in CT enclosures.
Note: For providing Code-required bending space at terminals for enclosed switches or individually enclosed circuit breakers, refer to 404.18. For conductor bending space at panelboard terminals, refer to 408.55. In Fig. 312-3, the clearances shown are determined from Table 312.6(A) or Table 312.6(B), under the column for one wire per terminal. For multiple-conductor circuit makeups, the clearance at terminals and inside gutters has to be greater, as shown under two, three, four, etc., wires per terminal.
Exception No. 2 of part (B) covers application of conductors entering or leaving a meter-socket enclosure, and was based on a study of 100- and 200-A meter sockets with lay-in terminals.
Fig. 312-3. These clearances are minimums that must be observed. (Sec. 312.6.)
Paragraph (C) applies to all conductors of size No. 4 or larger entering a cabinet or box from rigid metal conduit, flexible metal conduit, electrical metallic tubing, and so forth. As Indicated In 300.4(G), to protect the conductors from cutting or abrasion, a smoothly rounded insulating surface is required. While many fittings are provided with insulated sleeves or linings, it is also possible to use a separate insulating lining or sleeve to meet the requirements of the Code. Figure 312-4 shows use of a bushing with an insulated edge or a completely non-metallic bushing to satisfy this rule. Figure 312-5 shows an approved sleeve which may be used to separate the conductors from the raceway fitting, which may be installed after the conductors are already installed and connected.
Fig. 312-4. An insulated-throat bushing or other protection must be used at enclosure openings. (Sec. 312.5.)
Fig. 312-5. Slip-over nonmetallic sleeve may be used to cover metal bushing throat. (Sec. 312.5.)
In the exception to 300.4(G), an insulated throat is not required for conductor protection on enclosure threaded hubs or bosses that have a rounded or flared entry surface. This is recognition of a long-standing reality—that there is no need for protective insulating material around the interior opening of integral hubs and bosses on equipment enclosures. Insulated-throat bushings and connectors are needed only for entries through KOs in sheet-metal enclosures.
The last paragraph of 300.4(G) prohibits use of a plastic or phenolic bushing (“wholly of insulating material”) as a device for securing conduit to an enclosure wall. On a KO, there must be a metal locknut outside and a metal locknut inside to provide tight clamping to the enclosure wall, with the nonmetallic bushing put on after the inside locknut. An EMT or conduit connector must also be secured in position by a metal locknut and not by a nonmetallic bushing.
Because 300.4(G) is referenced, Part (C) also requires that any insulating bushing or insulating material used to protect conductors from abrasion must have a temperature rating at least equal to the temperature rating of the conductors.
312.8. Enclosures for Switches or Overcurrent Devices. The basic rule here is a follow-up to the rule of 312.7.
Most enclosures for switches and/or overcurrent devices have been designed to accommodate only those conductors intended to be connected to terminals within such enclosures. And in designing such equipment it would be virtually impossible for manufacturers to anticipate various types of “foreign” circuits, feed-through circuits, or numerous splices or taps.
The rule here states that enclosures for switches, CBs, panelboards, or other operating equipment must not be used as junction boxes, troughs, or raceways for conductors feeding through or tapping off, unless designs suitable for the purpose are employed to provide adequate space. This rule affects installations in which a number of branch circuits or subfeeder circuits are to be tapped from feeder conductors in an auxiliary gutter, using fused switches to provide disconnect and overcurrent protection for the branch or subfeeder circuits. It also applies to feeder taps in panelboard cabinets.
In general, the most satisfactory way to connect various enclosures together is through the use of properly sized wireways or auxiliary gutters (Fig. 312-6) or junction boxes. Figure 312-7 shows a hookup of three motor disconnects, using a junction box to make the feeder taps. Following this concept, enclosures for switches and/or overcurrent devices will not be overcrowded.
Fig. 312-6. Feeder taps in auxiliary gutter keep feeder cables and tap connectors out of switch enclosures. (Sec. 312.8.)
There are cases where large enclosures for switches and/or overcurrent devices will accommodate additional conductors, and this is generally where the 40 percent (conductor space) and 75 percent (splices or taps) at one cross section would apply. An example would be control circuits tapped off or extending through 200-A or larger fusible switches or CB enclosures. The csa within such enclosures is the free gutter wiring space intended for conductors.
Fig. 312-7. Junction box (arrow) is used for tapping feeder conductors to supply individual motor branch circuits—as shown in inset diagram. (Sec. 312.8.)
The last sentence in this rule is shown in Fig. 312-8 and is applied as follows:
example If an enclosure has a gutter space of 3 by 3 in., the csa would be 9 sq in. Thus, the total conductor fill (use Table 5, Chap. 9) at any cross section (including conductors) could not exceed 6.75 sq in. (9 × 0.75).
Fig. 312-8. These hookups are permitted where space in enclosure gutters satisfies the exception to the basic rule. (Sec. 312.8.)
In the case of large conductors, a splice other than a wire-to-wire “C” or “tube” splice would not be acceptable if the conductors at the cross section are near a 40 percent fill, because this would leave only a 35 percent space for the splice. Most splices for larger conductors with split-bolt connectors or similar types are usually twice the size of the conductors being spliced. Accordingly, where larger conductors are to be spliced within enclosures, the total conductor fill should not exceed 20 percent to allow for any bulky splice at a cross section.
Figure 312-9 shows an example of feeder taps made in panelboard side gutter where the cabinet is provided with adequate space for the large feeder conductors and for the bulk of the tap devices with their insulating tape wrap.
Fig. 312-9. In 312.11(C), the Code permits feeding through and tapping off in cabinets for panelboards on feeder risers, where the side gutter is specially oversized for the application. (Sec. 312.11.)
314.1 Scope. This rule makes clear that Art. 314 regulates use of conduit bodies when they are used for splicing, tapping, or pulling conductors. And this article does refer specifically to conduit bodies, to more effectively distinguish rules covering boxes, conduit bodies, and fittings. The rules of Art. 314 must be evaluated in accordance with the definitions given in Art. 100 for “conduit body” and “fitting.” Capped elbows and SE elbows are fittings, not conduit bodies, and must not contain splices, taps, or devices. The pieces of equipment described in Art. 314 tie into 300.15, “Boxes or fittings—where required.”
314.2. Round Boxes. The purpose of this rule is to require the use of rectangular or octagonal metal boxes having, at each knockout or opening, a flat bearing surface for the locknut or bushing or connector device to seat against a flat surface. But round outlet boxes may be used with nonmetallic-sheathed cable because the cable is brought into the box through a knockout, without the use of a box connector to secure the cable to the box. However, the exception to 314.17(C) permits only boxes of “nominal size 2¼ in. by 4 in.,” the so-called single gang boxes, to be used without securing the NM or NMC cable to the box itself—as long as it is stapled to the stud or joist within 8 in. (203 mm) of the box. Because round boxes are not single gang boxes, it appears that all such round outlet boxes must be equipped with cable clamps to satisfy the exception to 314.17(C) (Fig. 314-1). Shallow metal boxes with internal clamps for NM cable are acceptable as round boxes.
Fig. 314-1. Round boxes may be used only for connecting cables with internal clamps—such as NM or BX cable. (Sec. 314.2.)
314.3. Nonmetallic Boxes. Growth in the application of nonmetallic boxes over past years is the basis for the two exceptions to this section, which regulate the conditions under which nonmetallic boxes may be used with metal raceways or metal-sheathed cable. The need and popularity of these boxes developed out of industrial applications where corrosive environments dictated their use to resist the ravages of various punishing atmospheres. In many applications it is desirable to use nonmetallic boxes along with plastic-coated metal conduits for a total corrosion-resistant system. Such application is recognized by the Code in the exceptions of this section, although a limitation is placed requiring “internal” or “integral” bonding means in such boxes (Fig. 314-2).
According to the basic rule in the first sentence of this rule, nonmetallic boxes are permitted to be used only with open wiring on insulators, concealed knob-and-tube wiring, nonmetallic sheathed cable, electrical nonmetallic tubing, and rigid nonmetallic conduit (any “nonmetallic raceways”). Exception No. 1 requires internal bonding means in such boxes used with metal cable or raceways. The permission used to apply only to nonmetallic boxes sufficiently large—that is, over 100 cu in. Now, any size of box—PVC boxes, fiberglass boxes, or other nonmetallic boxes or enclosures—may be used with metal raceways or metal-sheathed cable. However, for each entry, a bonding bushing would have to be applied in order to provide continuity in the equipment grounding conductor.
Exception No. 2 requires that “integral” bonding means between all “threaded” raceway and cable entries must be provided in the box for all metal conduits or metal-jacketed cables. This exception addresses an entirely different box construction, where a metallic web is positioned within and throughout a nonmetallic (usually fiberglass) enclosure. Many of these boxes are used for hazardous locations that are also extremely corrosive. Some of them even meet the requirements for Class I Division 1. When hubs are added to these boxes using the methods prescribed in the installation instructions, continuity is achieved through the integral web, almost as if the box were all metal. The only constraint imposed in this case is that an equipment grounding terminal must be provided within the box by its manufacturer, which would reach the bonding web. That way if there is equipment in the box that requires a bonding connection, it will be available.
Fig. 314-2. Nonmetallic boxes are recognized for use with metal raceways and metal-sheathed cable. (Sec. 314.3.)
314.4. Metal Boxes. This is now a very simple rule. Ground all metal boxes. Period. The only exception is when 250.112(I) relieves the need for equipment grounding.
314.5. Short-Radius Conduit Bodies. For generations the conduit bodies listed here have been in a sort of “no-man’s land.” They meet the Art. 100 definition of a conduit body. They fail the bending space rules in 314.28(A)(2). They are obviously unsuited to enclose splices or devices. So now they have their own special classification, along with some rules to make sure they are applied properly. They must not be used with any conductors 4 AWG and larger, for which bending space rules apply, and they must only serve to change a raceway direction, nothing more. In addition to the items named in the rule, the classification includes flat 90°Fittings with the removable covers that are coplanar with the axis of the raceway entries, and also the little “handy ells” often used at the edge of a box.
314.15. Damp or Wet Locations. In damp or wet locations, equipment within the scope of Art. 314 must be placed or equipped to prevent moisture from entering or accumulating within the equipment. Boxes, conduit bodies, and fittings used in wet locations must be listed for use in those areas. The left part of Fig. 314-3 shows a fitting which is generally available as a listed item and it has been designed with the openings for the conductors so placed that rain or snow cannot enter the fitting. On the right, it shows a fitting also suitable for a wet location listing; it has a metal cover that slides under flanges on the face of the fitting, and, as required by 230.53, an opening is provided through which any moisture condensing in the conduit can drain out. Note that this is a “service entrance elbow.” It is a short-radius conduit body covered in 314.5. It is functionally obsolete for service applications because it does not have the required bending radius for any conductors 4 AWG or larger. It does retain theoretical acceptability for a service to a single small load such as a telephone booth, as covered in 230.42(B) and 230.79(A).
Fig. 314-3. Fittings must be suited to use in wet locations. (Sec. 314.15.)
314.16. Number of Conductors in Outlet, Device, and Junction Boxes, and Conduit Bodies. Note that motor terminal housings are excluded from the rules on box conductor fill. And where any box or conduit body contains 4 AWG or larger conductors, all the requirements of 314.28 on pull boxes must be satisfied. Refer to 314.28 for applications of conduit bodies as pull boxes.
Selection of any outlet or junction box for use in any electrical circuit work must take into consideration the maximum number of wires permitted in the box by 314.16. Safe electrical practice demands that wires not be jammed into boxes because of the possibility of nicks or other damage to insulation—posing the threat of grounds or shorts.
This section is broken down into three subparts. Part (A) establishes the volume of a box. Part (B) describes the method for determining how much volume is used by the various conductors, devices, and the like. Part (C) applies to conduit bodies, only.
As stated in part (A)(1) of this section, Table 314.16(A) shows the maximum number of wires permitted in the standard metal boxes listed in that table. But that table applies only where all wires in a given box are all of the same size; that is, all 14 AWG or all 12 AWG Table 314.16(B) is provided for sizing a box where all the wires in the box are not the same size, by using so much cubic-inch space for each size of wire.
Table 314.16(A) includes the maximum number of 18 AWG and 16 AWG conductors that may be used in various sizes of boxes, and Table 314.16(B) gives the required box space for those sizes of conductors. Because of the extensive use of 18 AWG and 16 AWG wires for fixture wires and for control, signal, and communications circuits, these data are needed to assure safe box fill for modern electrical systems.
As stated in part (A)(2), all other boxes—nonstandard, nonmetallic, or those metal boxes covered by part (A)(1) that are stamped with their cubic-inch capacity by the manufacturer—must consider their volume to be that which is stamped on them. And, as stated in the first paragraph of 314.16, the value of volume [part (A)] must never be less than the fill [part (B)].
Part (A)(2) of 314.16 covers boxes—metal and nonmetallic—that are not listed in Table 314.16(A). And the basic way of determining correct wire fill is to count wires in accordance with the intent of 314.16(B) and then calculate the required volume of the box or conduit body by totaling up the volumes for the various wires from Table 314.16(B).
Part (A)(2) covers wire fill for metal boxes, up to 1650 cu cm (100 cu in.) volume, that are not listed in Table 314.16(A) and for nonmetallic outlet and junction boxes. Although Code rules have long regulated the maximum number of conductors permitted in metal wiring boxes [such as given in Table 314.16(A)], there was no regulation on the use of conductors in nonmetallic device boxes up to the 1978 NEC. Since that time, 314.16(A)(2) requires that both metal boxes not listed in Table 314.16(A) and nonmetallic boxes be durably and legibly marked by their manufacturer with their cubic inch capacities to permit calculation of the maximum number of wires that the Code will permit in the box. Calculation of the conductor fill for these boxes will be based on the marked box volume and the method of counting conductors set forth in 314.16(B). The conductor volume will be taken at the values given in Table 314.16(B), and allocations of space as required for wiring devices or for clamps must be made in accordance with the rules of 314.16(B). This requirement for marking of both metal and nonmetallic boxes arises from the wording of 314.16(A)(2), which refers to boxes other than those described in Table 314.16(A) and to nonmetallic boxes.
As shown in Fig. 314-4, a nonmetallic box for a switch has two 14/2 NM cables, each with a 14 AWG ground. The wire count is four 14 AWG insulated wires, plus two for the switch to be installed, and one for the two ground wires. That is a total of seven 14 AWG wires. From Table 314.16(B), at least 32.8 cu cm (2 cu in.) of box volume must be allowed for each 14 AWG. This box must, therefore, be marked to show that it has a capacity of at least 7 × 2, or 14, cu in. As shown, the ground wires are connected by a twist-on connector, with one end of the wire brought out to connect to a ground screw on the switch mounting yoke. Such a technique is required to provide grounding of a metal switchplate. Refer to 404.9(B).
Fig. 314-4. Every nonmetallic box must be “durably and legibly marked by the manufacturer” with its cubic-inch capacity to permit calculation of number of wires permitted in the box—using Table 314.16(B) and the additions of wire space required to satisfy 314.16(A)(1). (Sec. 314.16.)
Part (B) of 314.16 describes the detailed way of counting wires in a box and subtracting from the permitted number of wires shown in Table 314.16(A) where cable clamps, fittings, or devices like switches or receptacles take up box space.
Important details of the wire-counting procedure of part (B) are as follows:
1. From the wording, it is clear that no matter how many ground wires come into a box, whether they are ground wires in NM cable or ground wires run in metal or nonmetallic raceways, a fill allowance of only one conductor must be made from the number of wires shown in Table 314.16(A) (Fig. 314-5). Or, as will be shown in later examples, one or more ground wires in a box must be counted as a single wire of the size of the largest ground wire in the box. Any wire running unbroken through a box counts as one wire, provided it is not long enough to be cut in the center of the loop and used as two wires, as defined in 300.14; if it is long enough to count as two wires, then it must be counted as two wires, and the count goes up further on multiple coils. A small hand loop that is only there to assist routing the wire is still only counted once. Each wire coming into a splice device (crimp or twist-on type) is counted as one wire. And each wire coming into the box and connecting to a wiring device terminal is one wire.
When a number of “isolated-ground” equipment grounding conductors for receptacles come into a box along with conventional equipment grounding wires, each type of equipment ground wires must be counted as one conductor for purposes of wire count when determining the maximum number of wires permitted in a box. When a number of isolated-ground receptacles are used in a box (as for computer wiring), all the isolated-ground conductors count as one fill allowance counted against the number of wires given in Table 314.16(A) as permitted for the particular size of box. And then, another allowance of one conductor must be made for any other equipment grounding conductors (not isolated-ground wires).
Fig. 314-5. Count all ground wires as one wire (or two wires if isolated-ground wires are also used) of the largest size of ground wire in the box. (Sec. 314.16.)
There is one type of wire entry that does not need to be counted, and that is fixture wires smaller than 14 AWG (maximum of 4 with an additional equipment grounding conductor permitted) that enter a box from a domed luminaire canopy or equivalent. This wording includes the canopies on paddle fan hanger assemblies. The word domed is there to assure at least some volume contribution from the canopy; there are luminaires that incorporate what for all intents and purposes is a 100-mm (4-in.) octagon blank cover; in such cases the fixture wires are counted. The domed canopy allowance makes it possible to hang luminaires on ceiling pans that usually have no more than 98.4 cm3 (6.0 in.3) for volume. Until comparatively recently wires smaller than 14 AWG had no mandatory wire fill assigned, however, now that those volumes are assigned, something had to be done to be certain that ceiling pans remained workable.
2. Regarding the addition of a wire fill allowances to other such allowances under consideration, for fixture studs, cable clamps, and hickeys, does this apply to the previously mentioned items collectively regardless of number and combination, or does it apply to each item individually—such as clamps, plus one; studs, plus one; and so forth?
Answer: It is the intent of parts (B)(2) and (3) to clarify that a fill allowance of one must be made from the number in the table for each type of device used in a box, and this must use the largest conductor in the box as the reference point for the calculation. A fill allowance of one must be made if the box contains cable clamps—whether one clamp or two clamps, a count of only one has to be made. A fill allowance of one must be made if the box contains a fixture stud. A fill allowance of one must be made if the box contains a hickey. Thus, a box containing two clamps but no fixture studs or hickeys would have a fill allowance of one from the table number of wires for the clamps. If a box contained one clamp and one fixture stud, a fill allowance of two would be made because there are two types of devices in the box.
Then, as given in part (B)(4), in addition to the allowances for clamps, hickeys, and/or studs, an allowance of two conductors must be made for each mounting strap that supports a receptacle, switch, or combination device. Note that for large devices that cannot fit in single-gang metal device boxes listed in Table 314.16(A), such as NEMA 14-50R (3-pole 4-wire 50-A 125/250-V) receptacles for ranges, allowances must be taken on the basis of the number of such ganged device boxes required for the device. In this case that would be two such boxes, for a total of four allowances, and if 6 AWG conductors are used, this rule adds another 163.8 cu cm (10 cu in.) to the volume.
3. Must unused cable clamps be removed from a box? And if clamps are not used at all in a box, must they be removed to avoid the one-wire fill allowance?
Answer: Unused cable clamps may be removed to gain space or fill in the box, or they may be left in the box if adequate space is available without the removal of the clamp or clamps. If one clamp is left, the one-wire allowance must be made. If no clamps are used at all in a box, such as where the cable is attached to the box by box connectors, the one-wire deduction is not made.
4. Is the short jumper installed between the grounding screw on a grounding-type receptacle and the box in which the receptacle is contained officially classified as a bonding jumper? And is this conductor counted when the box wire count is taken?
Answer: The jumper is classed as a bonding jumper. 250.146 uses the wording “bonding jumper” in the section pertaining to this subject.
However, whatever it is called, this conductor is not counted because it does not leave the box. The last sentence of 314.16(B)(1) covers that point.
The first sentence of 314.16(A) requires that ganged boxes be treated as a single box of volume equal to the sum of the volumes of the sections that are connected together to form the larger box. An example of wire counting and correct wire fill for ganged boxes is included in the following examples. Note: In the examples given here, the same rules apply to wires in boxes for any wiring method—conduit, EMT, BX, NM.
The top example in Fig. 314-6 shows how wire fill allowances must be added to account for a box containing cable clamps and a fixture stud. Note this is an example of a domed canopy and the two fixture wires from the luminaire are not be counted. The example at the bottom shows a nonmetallic-sheathed cable with three 14 AWG copper conductors supplying a 15-A duplex receptacle (one ungrounded conductor, one grounded conductor, and one “bare” grounding conductor).
After supplying the receptacle, these conductors are extended to other outlets and the conductor count would be as follows:
The 14 AWG conductor column of Table 314.16(A) indicates that a device box not less than 3 by 2 by 3½ in. (75 by 50 by 90 mm) is required. Where a square box with plaster ring is used, a minimum of 4- by 1¼-in. (100- by 32-mm) size is required.
Table 314.16(A) includes the most popular types of metal “trade-size” boxes used with wires 14 AWG to 6 AWG. Cubic-inch capacities are listed for each box shown in the table. According to paragraph (A)(2), boxes other than those shown in Table 314.16(A) are required to be marked with the cubic-inch content so wire combinations can be readily computed.
Figure 314-7 shows another example with the counting data in the caption. The wire fill in this case may violate the limit set by 314.16(A).
Figure 314-8 shows an example of wire-fill calculation for a number of ganged sections of sectional boxes. The photo shows a four-gang assembly of 3-by 2- by 3½-in. (75- by 50- by 90-mm) box sections with six 14/2 NM cables, each with a 14 AWG ground wire and one 14/3 NM cable with a 14 AWG ground. The feed to the box is 14/3 cable (at right side), with its black wire supplying the receptacle which will be installed in the right-hand section. The red wire serves as feed to three combination devices—one in each of the other sections—each device consisting of two switches on a single strap. When finished, the four-gang box will contain a total of six switches and one duplex receptacle. Each of the 14/2 cables will feed a switched load. All the white neutrals are spliced together and the seven bare 14 AWG ground wires are spliced together, with one bare wire brought out to the receptacle ground terminal and one to the ground clip on the bottom of the left-hand section. The four-gang assembly is taken as a box of volume equal to 4 times the volume of one 3- by 2- by 3½-in. (75- by 50- by 90-mm) box. From Table 314.16(A), that volume is 18 cu in. for each sectional box. Then, for the four-gang assembly, the volume of the resultant box is 4 × 18, or 72, cu in. This is the front half of the calculation, set it aside.
Fig. 314-6. Correct wire count determines proper minimum size of outlet box. (Sec. 314.16.)
Remember that the box is to be considered as one box, indivisible. Do not waste time trying to apportion volumes or fill allowances by gang. Do not make wire counts by Table 314.16(A) numbers per gang, because they are truncated. For example, the square box in the previous example holds 10 14 AWG wires, but this number is rounded down because the code limit is a not to exceed number. Its capacity is actually 344 cm3 (21.0 in.2); if these were gangable, and several were put together, and you figured on the basis of 10 wires each, you would lose one wire for every two boxes put together.
