600.1. Scope. In the case of signs that are constructed at a shop or factory and sent out complete and ready for erection, the inspection department must require listing and installation in conformance with the listing. In the case of outline lighting and signs that are constructed at the location where they are installed—the so-called skeleton tubing covered by part II—the inspection department must make a detailed inspection to make sure that all requirements of this article are complied with.
Rules governing the installation of electric signs vary widely from jurisdiction to jurisdiction. In some cities, inspection departments inspect signs in local shops as well as performing installation site inspections. Likewise, in some cities, the electrical inspector conducts the sign inspection alone. In others, a sign inspector will review plans for proposed signs for compliance with local ordinances, and the electrical inspector will then do the field inspection in conjunction with the sign inspector. Contact the building department in the municipality having jurisdiction to establish the exact procedure that must be followed.
600.2. Definitions. This section contains a number of definitions that are critical to the proper application of the Code rules for electric signs and outline lighting.
600.3. Listing. All electric signs, section signs, and outline lighting, whether fixed, mobile or portable, must be listed unless exempted by special permission. However, field-installed skeleton tubing is not required to be listed if installed per NEC rules, and outline lighting need not be listed as a system if it consists of listed luminaires wired using Chap. 3 methods.
600.5. Branch Circuits. In part (A), the Code mandates that any commercial occupancy that is “accessible to pedestrians” must be provided with one outlet for the purpose of supplying an electric sign, which must be accessible. Additionally, this section requires that the branch circuit supplying this outlet be a dedicated circuit, with no other loads supplied. The wording of this section requires that a sign outlet be installed for every ground-level store—even if an outdoor electric sign is not actually installed or planned. Note that no limit is placed on the number of outlets that may be connected on one circuit for a sign or for outline lighting, except that the total load should not exceed 16 A where incandescent or fluorescent lighting loads are to be supplied. A 30-A maximum is established for branch circuits supplying neon tubing. However, the minimum size for this circuit is 20 A. Note that for a mall-type environment, the wording is for “each entrance” so the outlet could be on the central hallway.
Where the loads to be supplied are “continuous loads,” that is, where in normal operation the load will continue for 3 h or more, the load should not exceed 80 percent of the branch-circuit rating. Given that commercial lighting is generally considered to be a “continuous load,” circuits and OC devices used to supply and protect such loads must be sized on the basis described for continuous loading. That is, the conductors must have current-carrying capacity that is at least 125 percent of the continuous load, before derating. And the OC device must also have a long-time trip rating of at least 125 percent of the rated lighting load. This effectively translates into a 16-A maximum lighting load and No. 12 copper conductors for a 20-A circuit, and 24 A using No. 10 copper on a 30-A circuit. See 210.19(A), 210.20(A), and 240.4(D).
Part (B) indicates the minimum ratings for the required sign circuit. The 20-A branch circuit for sign and/or outline lighting for commercial occupancies with ground-floor pedestrian entry—required by part (A)—may supply one or more outlets for the purpose, but not any other loads. The intent is that the required, dedicated 20-A circuit supply one or more outlets intended for electric signs. However, if the intended sign is to be made of neon tubing, the rating could be 30 A (and no more).
As noted in part (C), the wiring method used to supply signs must conform with the requirements of parts (1), (2), and (3).
In part (C)(1), the Code requires that the wiring method used to supply the sign—which may be any of the wiring methods recognized in Chap. 3 suitable for the type of location in which the sign is installed—must be terminated either in the sign, in a box provided with the sign, or in a typical junction/outlet box or conduit body. In part (C)(2), the Code recognizes the use of signs and electric-sign transformer enclosures as raceways to supply adjacent signs and associated equipment. And part (C)(3) mandates that metal poles used to support electric signs comply with the rules for poles used to support lighting, as covered in 410.30(B). (See Fig. 600-1).
600.6. Disconnects. Figure 600-2 depicts the disconnecting means that should be within sight of the sign, outline lighting, or remote controller, as covered in part (A)(1). The phrase “within sight” is clearly defined in Art. 100, and it is well understood that it means the same thing as the phrase “in sight from,” which specifies that it shall be visible and not more than 50 ft (15.0 m) distant from the other. Some signs are comprised of sections as part of a listed unit (see the “Section Sign” definition in 600.2) and those sections may in some cases be placed where they continue around the corner of a building. In such cases part of the sign will be “out of the line of sight” from other sections, and the disconnect must be capable of being locked in the open position. Note that section signs must be marked to indicate that field wiring and installation instructions are required.
Fig. 600-1. Commercial buildings must have outdoor sign outlet. (Sec. 600.5.)
Fig. 600-2. An “in-sight” disconnect may be in the sign or visible from the sign. (Sec. 600.6.)
Figure 600-3 illustrates the conditions recognized by part (A)(2), which allows the disconnecting means to be located within sight of the controller where the signs are operated by electronic or electromechanical controllers located external to the sign. Note that any sign disconnect located within sight of the controller must be capable of being locked in the “off” or “open” position, as specified in part (A)(3).
Fig. 600-3. Controller disconnect location may vary, but disconnect must be lock-open type. (Sec. 600.6.)
With respect to part (B), any switching device controlling the primary of a transformer that supplies a luminous gas tube operates under unusually severe conditions. In order to avoid rapid deterioration of the switch or flasher due to arcing at the contacts, the device must be rated for inductive loads or have a current rating of at least twice the current rating of the transformer it controls. General use ac snap switches, as covered in 404.14, are rated for inductive loads and will handle motor loads up to 80 percent of their rating, making then suitable for this application provided a permanent provision has been made for locking them open. There are handy-box and square-box raised covers available with factory mounted escutcheons around the toggle slot that will accept padlocks.
The question of what constitutes “operated by electronic or electromagnetic controllers” frequently arises in the context of modern energy management systems that turn signs on and off along with other building lighting. The purpose of the disconnect rule is to provide maintenance personnel with a secure means to ensure that the equipment they are working on is disconnected and that it will stay that way until they are ready to reenergize it. By long usage and custom, a maintenance disconnect is unique to its equipment. No one would seriously suggest that because a service disconnect could be locked in the open position, all NEC requirements for disconnects of specific loads were met. Any maintenance worker on any downstream equipment would feel compelled to work the downstream equipment hot rather than inconvenience the enterprise to that extent. Therefore, if the energy management system operates a contactor for a sign among other control devices for other loads, this section requires that contactor to be in sight of a disconnect, that disconnect will be capable of being locked open, and a second disconnect would not be required at the sign location. On the other hand, if the contactor operated a lighting panel for which the sign was one load among many, then a local disconnect for the sign must be installed.
600.7. Grounding and Bonding. The general principles in this section agree with the grounding principles in Art. 250, but there are specific provisions that are unique to this article, particularly with respect to neon signs that operate at high voltage and very low current. Part (A) addresses equipment grounding connections to sign parts that may be exposed to line voltage and follow the usual rules for equipment grounding connections. Specific language forbids the use of metal building parts as an equipment grounding return path, reiterating 250.136(A). Small metal parts, such as the metal feet on tubing support clips, do not require bonding connections.
Part (B) addresses bonding, and addresses neon signage with provisions that are unique throughout the NEC. Flexible metal conduit is permitted in total accumulated lengths of up to 30 m (100 ft) as a bonding conductor, although the run from the transformer to the first termination will not exceed 6 m (20 ft) per 680.32(J)(2). The currents are very small and the performance of this wiring method in this context was investigated by a testing laboratory. If the wiring to the sign is in a nonmetallic wiring method such as PVC conduit, the bonding conductor must run outside the conduit, with the spacing governed by the frequency of the power supply (Fig. 600-4). For transformers with no change in frequency the spacing from conductor to conduit must be at least 38 mm (1½in.) and for electronic power supplies with frequencies over 100 Hz the spacing must be at least 45 mm (1¾ in.). This reduces the voltage gradient between the GTO cable and the bonding conductor, and in so doing reduces the likelihood of damaging corona discharge. Metal raceways distribute the ground plane equally around the cable, and are also shorter, as noted in Fig. 600-4. Research showed that most of the damage occurs in the initial cable run from the power supply.
The NEC does address, courtesy (originally) of Las Vegas, the wiring of signs in fountains. Most of the requirements are in 680.57, covered later in this chapter. However the bonding requirements are here, in that the metal piping system, required to be bonded by 680.53, is permitted as the bonding connection required here. If the piping is nonmetallic, then a connection must be arranged to the equipment grounding conductor.
Fig. 600-4. Many unique bonding requirements apply to neon installations. (Sec. 600.7.)
600.10. Portable or Mobile Signs. This rule applies to outdoor portable or mobile signs that are plug-connected. The rule calls for “factory-installed” GFCI protection in or within 12 in. (300 mm) of the plug cap at the end of the supply cord from the sign to protect personnel from potential shock hazards. Documentation for the need for this new rule cited six accidents—three deaths and three shocks—due to ground faults in such outdoor signs that were plug-connected but in which there was no grounding connection or a failed grounding connection. Note that the product standard requires that this integral protection include open-neutral protection, as discussed in the previous chapter (590.6).
600.21. Ballasts, Transformers, and Electronic Power Supplies. The transformers used to supply luminous gas tubes are, in general, constant-current devices and, up to a certain limit, the voltage delivered by the transformer increases as the impedance of the load increases. The impedance of the tube increases as the length increases and is higher for a tube of small diameter than for one of larger diameter. Hence, a transformer should be selected which is designed to deliver the proper current and voltage for the tube. If the tube is too long or of too small a diameter, the voltage of the transformer may rise to too high a value. The maximum voltage permitted is center-grounded 15 kV (max. voltage to ground on each leg 7500 V) with 300 mA as the maximum current allowed, as covered in 600.23(C&D).
This section includes numerous rules on location of and access to this equipment. The first consideration is to minimize the length of the secondary conductors, which places the power supply as close as possible to the sign or outline lighting. Locations above suspended ceilings are acceptable as long as the branch-circuit connections use Chap. 3 wiring methods and not flexible cord. Wherever located, the equipment must have workspace for servicing, defined as a cubical volume 1 yd (900 mm) on a side, and if located in an attic, there must be access to this space through an access door no thinner than the spacing between standard framing on 24-in. (610 mm) centers, and at least 900 mm (3 ft) high. There must be a permanent walkway to the location at least 300 mm (1 ft) wide and the workspace must be lit with the lighting controlled either by a snap switch or a pull chain, as long as the control can be activated at the point of entry.
600.23. Transformers and Electronic Power Supplies. This equipment must be listed, and must incorporate, with limited exceptions, secondary-circuit ground-fault protection. Arcing failures on the secondary side of these transformers involve very small but ignition-capable currents. This major design innovation responds to numerous documented fires from this source. The power supply must be marked accordingly. Note that the marking requirement is misworded, since it literally applies to all power supplies whether or not they actually have this protection; and some supplies that meet the integral containment or limited voltage provisions will not have it.
600.24. Class 2 Power Sources. Thanks to the evolution of LED light sources, some signs and outline lighting actually can operate under Class 2 power limitations, and those are covered here.
600.32. Neon Secondary-Circuit Wiring, over 1000 V, Nominal. These are the wiring rules governing the secondary side of a neon-lighting power supply. Most Chap. 3 raceways are permitted, but ENT is not, based on concerns it may sag over time in a way that would not have been apparent during the original installation or inspection, and drop the contained single-conductor GTO cable too close to a grounded surface. As noted in Fig. 600-4, these surfaces must also meet the distances required of bonding conductors, and for the same reasons. In running nonmetallic raceways, nonmetallic methods must be used to secure the raceways away from grounded building elements to the extent required. This rule does not, however, prevent the use of grounded metal raceways, since the ground plane evenly surrounds the cable. The final choice of wiring method involves consideration of permitted length of cable run (advantage: nonmetallic) versus spacing requirements (advantage: metallic) along with any other design factors. These may include recommendations of the power supply manufacturer, which may include limitations based on the amount of capacitive reactance his power supply can cope with. Note that the maximum distances specified here are not accumulative distances as in 600.7, but only to the first tubing connection. Where the GTO cable emerges from a metallic raceway, the insulated cable must extend not less than 65 mm (2½ in.) beyond the raceway end.
600.41. Neon Tubing. Part (D) requires protection in the form of guards or other approved means if the skeleton tubing is accessible to other than qualified persons. Also note that 600.32(I) forbids the use of this type of installation in (or on) a dwelling occupancy.
600.42. Electrode Connections. These requirements have been toughened over the years, and now require listed components for most of this part of the installation.
604.1. Scope. This article covers modular prefab wiring systems for ceiling spaces, raised floors, and increasingly walls as well.
These manufactured wiring systems were logically dictated by a variety of needs in electrical systems for commercial-institutional occupancies. In the interest of giving the public a better way at a better price, a number of manufacturers developed basic wiring systems to provide plug-and-receptacle interconnection of branch-circuit wires to luminaires in suspended-ceiling spaces. Such systems afford ready connection between the hard-wired circuit homerun and cables and/or ducts that form a grid- or tree-like layout of circuiting to supply incandescent, fluorescent, or HID luminaries in the ceiling.
Acknowledged advantages of modular wiring systems are numerous and significant:
Factory-prewired raceways and cables provide highly flexible and accessible plug-in connection to multicircuit runs of 120- and/or 227-V conductors.
Drastic reductions can be made in conventional pipe-and-wire hookups of individual circuits, which are costly and inflexible.
Plug receptacles afford a multiplicity of connection points for luminaires to satisfy needs for specific types and locations of lighting units to serve any initial layout of desks or other workstations while still offering unlimited, easy, and extremely economical changes or additions of luminaires for any future rearrangements of office landscaping or activities.
Systems may also supply switches and/or convenience receptacles in walls or partitions, with readily altered switching provisions to provide energy conservation through effective ON-OFF control of any revised lighting layout.
Work on the systems has been covered by agreement between the IBEW and associated trades.
Such systems have a potential for a tax advantage of accelerated depreciation as office equipment rather than real estate.
604.4. Uses Permitted. Modular systems may be used in air-handling ceilings. Equipment may be used in the specific applications and environments for which it is listed by UL. Note that one end of such a system can be concealed, such as when it is fished down a wall that is open above a suspended ceiling, and used to connect either a switch or an outlet.
604.6. Construction. Prewired plug-in connections may be AC or MC cable or metal flex. The permitted conductors must be copper, and limited to a minimum of 12 AWG and running up to 8 AWG, and including a fully sized copper equipment grounding conductor (insulated or bare) is always required in each cable or flex length—even though flex itself is otherwise permitted to be used by the NE Code without an equipment grounding conductor in lengths not over 6 ft (1.8 m), provided the wires within it are protected at not over 20 A, and AC cable in other uses is recognized by the Code and by UL for equipment grounding through its armor and enclosed aluminum bonding wire (Fig. 604-1). Type MC cable of the smooth or corrugated variety with a qualified armor and equipment grounding conductor combination is also permitted, provided it is equivalent to the ungrounded conductor sizing. The purpose for the 8 AWG is the reduction of voltage drop and not high ampere loading, and therefore the equivalent sizing is appropriate in these branch circuit sizes to meet 250.122(B).
The same sizing applies to conductors in flexible metal raceways. Here, however, a whip to a luminaire is permitted at not over 1.8 m (6 ft) using not less than 18 AWG taps. In addition, 12 AWG or larger flexible cord suitable for hard usage, not over 1.8 m (6 ft) long and visible over its entire length, is permitted as a supply to equipment not fixed in place, such as some display cases. This wiring system is also permitted to contain signaling and communications wiring, within the limits set in Art. 725 and Chap. 8.
Fig. 604-1. Modular wiring systems are fully recognized by the Code. (Sec. 604.6.)
605.1. Scope. This Code article covers electrical equipment that is part of manufactured partitions used for subdividing office space, as shown in Fig. 605-1.
605.2. General. Only those wiring systems “identified” to supply lighting and appliances may do so. Check the listing data and manufacturer’s installation instructions to ensure proper use and installation. The partitions must not extend from floor to ceiling without the permission of the inspector, and if they do reach the ceiling, they must not go above it in any way.
605.4. Partition Interconnections. Wired partitions may be interconnected by a cord and plug. If cord is used, the partitions must be “mechanically contiguous” and the cord must be of extra hard usage and of 12 AWG minimum, and no longer than required to make the connection and never longer than 600 mm (2 ft). The basic rule calls for interconnection of partitions by a “flexible assembly identified for use with wired partitions.”
605.8. Freestanding-Type Partitions, Cord-and-Plug-Connected. A partition or group of connected partitions that is supplied by cord-and-plug connection to the building electrical system must not be wired with multiwire circuits (all wiring must be 2-wire circuits) and not more than thirteen 15-A, 125-V receptacles may be used. The supply receptacle(s) must be on its (their) own circuit(s) and not more than 300 mm (12 in.) from the partition supplied. The supply cord must meet the same requirements as for interconnecting cords as covered in 605.4. Note that 605.6 requires fixed partitions to be connected using Chap. 3 wiring methods, and 605.7 does allow multiwire circuits to supply freestanding partitions, but only where Chap. 3 wiring methods bring them (must comply with 210.4) to the partition.
Fig. 605-1. This article covers electrical wiring and electrical components within or attached to manufactured partitions, desks, cabinets, and other equipment that constitute “office furnishings.” Photo at top shows interior wiring in base of partitions, to supply luminaires and receptacle outlets—as shown at arrows in bottom photo of a typical electrified office work station.
610.11. Wiring Method. In general, the wiring on a crane or a hoist should be raceways, or Type AC with a grounding conductor, Type MC or Type MI. However, for practical considerations, short lengths of flexible conduit or metal-clad cable and even open conductors may be used for connections to motors, brake magnets, or other devices where a rigid connection is impracticable because the devices are subject to some movement with respect to the bases to which they are attached. In outdoor or wet locations, liquidtight flexible metal conduit should be used for flexible connections.
610.14. Rating and Size of Conductors. Crane conductors operate for very limited time intervals, and the NEC includes a special ampacity table for this purpose. Note that there is exactly one place in the NEC where 5 AWG wire has formal recognition and it here in this table. Part (E) covers motor calculations such as where multiple motors could operate at one time. As covered here using nameplate data, take the largest motor or group of motors for any particular crane motion, and add 50 percent of the next largest motor or group of motors, in all cases using the column of Table 610.14(A) that applies to the longest time-rated motor.
610.21. Installation of Contact Conductors. Part (F) permits use of the track as one of the circuit conductors. In some cases, particularly where a monorail crane or conveyor is used for handling light loads, for the sake of convenience and simplicity it may be desirable to use the track as one conductor of a 3-phase system. Where this arrangement is used, the power must be supplied through a transformer or bank of transformers so that there will be no electrical connection between the primary power supply and the crane circuit, as in Fig. 610-1. The secondary voltage would usually be 220 V, and the primary of the transformer would usually be connected to the power distribution system of the building or plant. The leg connected to the track must be grounded at the transformer only, except as permitted in 610.21(F)(4).
Fig. 610-1. Isolating transformer is used to power track of crane or conveyor. (Sec. 610.21.)
610.32. Disconnecting Means for Cranes and Monorail Hoists. This disconnect is an emergency device provided for use in case trouble develops in any of the electrical equipment on the crane or monorail hoist, or to permit maintenance work to be done safely. A motor-circuit switch, molded-case switch, or circuit breaker must be provided in the leads from the runway contact conductors or other power supply on all cranes and monorail hoists. The disconnecting means must be capable of being locked in the open position. As in other places where the NEC discusses required disconnecting means, the provision for locking or adding a lock to the disconnecting means must be installed on or at the switch or circuit breaker used as the disconnecting means and must remain in place with or without the lock installed. Portable means for adding a lock to the switch or circuit breaker is not permitted.
Where a monorail hoist or hand-propelled crane bridge installation meets all of the following, the disconnecting means need not be installed:
1. The unit is controlled from the ground or floor level.
2. The unit is within view of the power supply disconnecting means.
3. No fixed work platform has been provided for servicing the unit.
Where the disconnecting means is not readily accessible from the crane or monorail hoist operating station, means must be provided at the operating station to open the power circuit to all motors of the crane or monorail hoist.
610.33. Rating of Disconnecting Means. It is possible that all the motors on a crane might be in operation at one time, but this condition would continue for only a very short while. A switch or CB having a current rating not less than 50 percent of the sum of full-load current rating of all the motors will have ample capacity. The continuous ampere rating of the switch or circuit breaker specified above shall be not less than 50 percent of the combined short-time ampere ratings of the motors nor less than 75 percent of the short-time ampere rating of the motors required for any single motion.
Note that Art. 610 uses the wording “within view” in several places instead of “in sight” consciously in order to avoid the 15 m (50 ft) limitation built into the NEC definition of “in sight.” Many large industrial cranes are to too big to make the 15 m (50 ft) limit workable.
610.61. Grounding. All exposed non-current-carrying metal parts of cranes and hoists must be bonded together, usually by the usual mechanical connections, to make an effective ground-fault current path. Moving parts may have their equipment grounding continuity established through metal-to-metal contact on bearing surfaces, making it unnecessary to run very long, strain-relieved bonding conductors that would need to accommodate major frame movements of perhaps hundreds of feet at large industrial facilities. However, as of the 2005 NEC, the contact between the wheels of a trolley (the part that makes the load go up and down) and its associated bridge girder can no longer be depended upon for grounding continuity, and now a bonding conductor must be installed between these two parts. The wording change only covers the trolley frame, and not the bridge girder as its wheels turn on the runway, even though the contact surfaces seem identical. There was no substantiation to distinguish one from the other, nor was there any loss experience presented to suggest that the prior allowance, unchanged since the 1962 NEC, was deficient.
620.1. Scope. These provisions may also be considered as applying to console lifts, equipment for raising and lowering or rotating portions of theater stages, and all similar equipment.
620.11. Insulation of Conductors. The major concern in this section is with the integrity of electrical systems in a hoistway that is a natural chimney in the event of fire. The hoistway door interlock wiring must be suitable for 200°C for this reason, and the traveling cables must be one of the kinds listed in Table 400.4 for this purpose. Other insulation must be flame retardant.
620.12. Minimum Size of Conductors. Code Tables 310.16 to 310.19 do not include the ampacity for 20 AWG copper conductors. However, it is generally considered that 20 AWG conductors up to two conductors in cable or cord may safely carry 3 A. This section amends 310.4 and permits those smaller conductors to be paralleled to equal the capacity of a 14 AWG wire in this case. This type of allowance makes for more flexible traveling cable make-ups, which is essential in today’s very tall buildings.
Because of wider use of advanced semiconductor computer equipment, use of wire smaller than 20 AWG is permitted for other than lighting circuits by part (B), with 24 AWG as the minimum, and even smaller if listed.
The development of elevator control equipment, which has been taking place for many years, has resulted in the design and use of equipment, including electronic unit contactors, requiring very much smaller currents (milliamperes) for their operation.
620.13. Feeder and Branch Circuit Conductors. This section tracks the requirements in Art. 430 quite closely, but throws in a demand factor table that is unique to multicar elevator groups, in Sec. 620.14. There are two examples in Annex D (Examples D9 and D10) that demonstrate multiple elevator loading calculations as covered in these code sections very well.
620.22. Branch Circuits for Car Lighting, Receptacle(s), Ventilation, Heating, and Air Conditioning. This section establishes that each elevator car will be served by two dedicated circuits, one for heating and air conditioning, and one for lighting, the receptacle on the cab top or similar for service work, and other accessory loads. The wiring sequence on this circuit must be such that the service receptacle, which will be a GFCI receptacle per 620.85, must not disconnect the cab light if it trips. The overcurrent protective devices for these circuits must be located in the machine room or other control space for that car. A similar requirement governs lighting and receptacles in machine rooms and similar control spaces, as given in the next section (620.23), and another similar requirement governs hoistway pit lighting and its required receptacle outlet as covered in the section after that (620.24).
620.32. Metal Wireways and Nonmetallic Wireways. This section allows the wire fill in wireways to more than double, from the 20 percent allowed in 376.22(A) to 50 percent. Note, however, that no mention is made of the derating factors in 310.15(B)(2)(a) that apply to the entire fill as soon as the 30 current-carrying conductors threshold is reached, and to all conductors in a nonmetallic wire-way. Since no exception to this Chap. 3 rule is taken in this Chap. 6 article, the derating factors will apply, and they could easily hit 0.35 (example: for 41 wires, 12 AWG THHN, 30 A ×0.35 = 10.5 A) if the wireway were actually stuffed to 50 percent fill.
620.36. Different Systems in One Raceway or Traveling Cable. It would be difficult, if not practically impossible, to keep the wires of each system completely isolated from the wires of every other system in the case of elevator control and signal circuits. Hence, such wires may be run in the same conduits and cables if all wires are insulated for the highest voltage used and if all live parts of apparatus are insulated from ground for the highest voltage, provided that the signal system is an integral part of the elevator wiring system. These are very sophisticated, listed cables, and if multiple cables were required, they could easily tangle in the long lengths that are required in today’s high rise applications.
620.37. Wiring in Hoistways, Machine Rooms, Control Rooms, Machinery Spaces, and Control Spaces. Although hoistways are very tempting chases to run from floor to floor, they are not permitted for this purpose. The only wiring permitted in an elevator hoistway is wiring for the elevator functions. Neither are hoistways permitted to contain, nor the vertical elevator rails permitted to be, down conductors for a lightning protection system. However, if a hoistway happens to run close enough to a down conductor for a lightning protection system such that NFPA 780 requires that rails in the hoistway be bonded to the down conductor to eliminate the possibility of side flash, then that bonding may proceed.
620.51Disconnecting Means. A disconnecting means must be provided for every elevator, and it must open all loads connected with that elevator except three, which are forbidden to be disconnected as a result of this disconnect being in the open position. The three exceptions are the cab lighting and accessories circuit, the cab heating and air conditioning circuit, and the hoistway and machine room lighting and service receptacle circuit(s). These must stay on so cab occupants will stay comfortable and not panic in the event of a malfunction, and so maintenance can proceed on the failed elevator under emergency conditions.
This main disconnecting means can be a switch or a circuit breaker. It can open automatically but must only close manually. It cannot be arranged to open from “any other part of the premises.” The adjacent hoistway is not considered an “other part of the premises.” Further, if sprinklers are installed in the hoistway, this disconnect must open prior to waterflow. The usual protocol to bring this together involves a smoke detector in the hoistway that will detect smoke considerably prior to when the sprinkler head fuses. Upon detecting smoke, the smoke detector, supervised as part of a fire alarm system, initiates a control sequence that parks the elevator on the recall floor and discharges passengers. Adjacent to the sprinkler head, a heat detector, also supervised by the fire alarm system, activates the shunt trip on the disconnect if the temperature continues to rise, doing so before the sprinkler head fuses to start the flow of water. This prevents waterflow onto an active elevator, which would be an extreme hazard as the brakes could slip and the controls fail.
620.53. Car Light, Receptacle(s), and Ventilation Disconnecting Means. There must be a disconnecting means for this circuit, described in 620.22, located in the machine room or control space for the associated elevator. The disconnecting means must be permanently capable of being locked in the open position. A factory-designed locking hasp for the circuit breaker originating this dedicated circuit would allow the circuit breaker to fulfill both requirements, provided it is in the machine room or control space for the associated car. It must be plainly labeled as to its function and the identity of the elevator for which it provides this function. As in the case of the required branch circuit for heating and air conditioning, the next section (620.54) provides parallel requirements for the heating and air conditioning circuit.
620.61. Overcurrent Protection. These requirements directly track other sections in the NEC.
620.62. Selective Coordination. This was the first mandatory selective coordination rule in the history of the NEC. It was substantiated on the basis that if an elevator feeder opened because of a fault, and if the upstream overcurrent protection opened due to a lack of coordination, people would be trapped in all the elevators stranded by the upstream protection opening, and service personnel might not be able to figure out on a timely basis how to find and reclose the upstream protection. Of course a sign on the elevator disconnect could solve that problem, but the panel insisted on solving a problem in a way that was both unprecedented and poorly substantiated. Depending on the relative size of the two levels of overcurrent protection and the trip curves for each this may be a simple or an intractable problem. Refer to the discussion at 700.27 for more information on a very controversial topic.
620.85. Ground-Fault Circuit-Interrupter Protection for Personnel. All 125-V 15- and 20-A receptacle outlets in pits, hoistways, elevator car tops, and in escalator and moving-walk wellways must be provided with GFCI protection, and this protection must be at the point of use, that is, through the use of a GFCI receptacle. A GFCI circuit breaker ahead of the receptacle is not permitted. The issue is the difficulty in resetting a tripped GFCI protective device that is nowhere near the worker who may need to reset the device immediately. On the other hand, the receptacle outlets in the machine room or in the machinery space can more easily reach the originating panelboard and rest the device, so either approach is permitted in those areas.
625.1. This article regulates the sizing and installation of equipment and conductors used to supply electrical vehicle charging systems and the electrical vehicle. This article does not apply to battery charging systems that are used for fork-lifts, etc., but rather automobile-type electrical vehicles.
625.9. Electric Vehicle Coupler. This is the interface between the premises wiring system and the vehicle electrical system for propulsion, and it must be polarized unless otherwise listed. The couplers must be configured to avoid inadvertent contact by untrained persons with uninsulated live parts. They cannot use a conventional NEMA plug and receptacle configuration, to avoid any confusion on the part of the public. Grounded and nongrounded configurations must be noninterchangeable as well. The coupling from connector to vehicle inlet must be positive so as to prevent unintentional disconnection. Unless listed otherwise, they must have a grounding pole that is first make/last break. Presumably the inductive paddles qualify as listed otherwise (see 625.16).
625.13. Electric Vehicle Supply Equipment. This equipment can be cord-and-plug connected if portable and rated not over 15 or 20 A at 125 V single phase. Otherwise it must be permanently connected and have no exposed live parts.
625.14. Rating. This must be high enough for the load to be served, and must be considered as continuous. Although some highway quick-charge protocols assume a 10- or 15-min recharge at very high ampere values, the rule is for a continuous classification on any load. The more usual anticipated charging protocol involves 32 A charging current from a 208-V wye or delta (or even single-phase) connection on a 40 A branch circuit.
625.15. Marking. In addition to an electric vehicle usage label, there will also be a marking with respect to whether or not ventilation is required. The vehicle owner should know which type of battery he has. Some are sealed or of a chemistry that does not emit hydrogen, and others do emit hydrogen. As provided in 625.29(C) if a vehicle requiring ventilation to remove hydrogen gas accumulations plugs into a charger marked ventilation not required, the charging station will listen for a coded signal indicating the appropriate battery is present, and not receiving it, will not charge the battery.
625.16. Means of Coupling. Again, the coupling can be inductive or conductive, and the connecting devices must be listed for the purpose.
625.17. Cable. Special cords have been developed for these connections, easily identifiable by the “EV” (electric vehicle) letters in their type designation. They can be either hard usage or extra-hard usage. They must not exceed 25 ft, and they can use the higher power cable ampacities in Table 400-5B (unless they are No. 10 or smaller). The rule also allows for hybrid cables that include signaling circuits (or optical fiber cables).
625.18. Interlock. The charging system must be designed with an interlock that de-energizes the cable and connector whenever it isn’t connected to the vehicle. This is not required on the 15- and 20-A 125 V charging systems. This inter-locking system will also prevent charging a vehicle that uses batteries that outgas hydrogen unless the installation is arranged with ventilation as provided in 625.29(D).
625.19. Automatic De-energization of Cable. The electric vehicle supply cord and system must include some method of de-energization in the event of excessive strain on the cable, such as by driving away while plugged in. This is not required on the 15- and 20-A 125 V charging systems,
625.21. Overcurrent Protection. As noted in the discussion under Sec. 625-14, these chargers are defined as a continuous load, and therefore the conductor ampacity and the overcurrent device must be increased by an additional 25 percent.
625.22. Personnel Protection System. The electric vehicle supply system must incorporate shock protection that may differ somewhat in trip thresholds from conventional GFCI levels, but that has the same effect. The test labs and manufacturers are being given some needed design flexibility here given that the output current may be dc or at different voltages. If the charging equipment is cord-and plug-connected, then this function must be built into the attachment plug or into the supply cord within a foot of the plug.
625.23. Disconnecting Means. High-capacity charging equipment (over 60 A or over 150 V to ground) must have a disconnecting means in a readily accessible location. It must be able to be locked in the open position. This is a disconnecting means for the equipment, and therefore a unit switch in the equipment would not comply, even if it opened all ungrounded conductors. Maintenance personnel must be free to service the entire unit without hazard.
Note that this equipment would meet the definition of an appliance (other than industrial, produced in standard sizes, etc.) and therefore must comply with 422.31(B). That rule also requires a local disconnect, which can only be the branch-circuit protective device if it is within sight or can be locked open. These provisions can be enforced on any capacity charging system.
625.25. Loss of Primary Source. The charging equipment must be arranged so the energy stored in the car batteries cannot backfeed into the supply wiring if the supply power fails. The vehicle cannot be allowed to serve, even if so desired, as a standby power source unless (as covered in 625.26) listed for this purpose, in which case the provisions in Art. 702 must be met for standby power and Art. 705 for interactive power, as applicable, must be met.
625.28. Hazardous (Classified) Locations. For charging equipment in these areas, the rules in Arts. 500-516 apply. This section needs to be here because Chaps. 5 and 6 have equal rank under 90.3. For example, 511.10(B)(1) requires extra-hard usage cord for this equipment in those service areas, thereby disallowing three of the six cord types otherwise itemized as acceptable in Sec. 625-17. This section removes the conflict.
625.29. Indoor Sites. These sites include both attached and detached residential garages, enclosed and underground parking “structures,” agricultural buildings, etc. The charging equipment must be located so the charging cable can connect directly to the vehicle. Unless listed differently, the vehicle coupling means must be stored or located within a zone between 450 mm and 1.2 m (18 in. and 48 in.) above the parking surface. Part (C) covers instances where ventilation is not required because of the nature of the battery system in the vehicle, as noted in 625.15(B); mechanical ventilation is not required in such cases.
One of the key aspects of this article is in the area of ventilation. Mechanical ventilation as covered in Part (D) must be provided with systems that are suitable for charging electric vehicles that outgas hydrogen and that are identified accordingly in 625.15(C). It must be permanently installed and it must include both supply and exhaust equipment arranged to take air from and exhaust air directly out to the outdoors. This ventilation must be interlocked with the charging system and it must operate during the entire charging cycle. The required volume of air to be exchanged is provided in a table based on the ampere rating and voltage of the branch circuit supplying the charging equipment. For example, a 20-A 120-V supply requires 49 cfm, and a 200-A 480-V 3-phase supply requires 3416 cfm. These numbers apply for each space equipped to charge an electric vehicle. If there are two spaces, then the required ventilation doubles.
A large proportion of battery charging involves the release of hydrogen gas. This is a Class I Group B gas, extremely dangerous, and its lower explosive limit is only 4 percent. That means that a hydrogen-air mixture over 4 percent hydrogen by volume can explode. Although hydrogen is much less dense than air and dissipates rapidly, charging operations can generate enormous quantities. Actual testing showed ignitable concentrations of hydrogen near the ceiling even on 15-A branch circuits in residential garages with the door open! The mechanical ventilation requirements in this section need to be taken seriously.
625.30. Outdoor Sites. This includes carports, open parking structures, curb-side units, etc. The charging equipment must be located so the charging cable can connect directly to the vehicle. Unless listed differently, the vehicle coupling must be stored or located within a zone between 600 mm and 1.2 m (24 and 48 in.) above the parking surface. Note this minimum height is greater than for indoor applications.
626.1. Scope. This article, new in the 2008 NEC, covers what is defined in 626.2 as a “truck parking space that has been provided with an electrical system that allows truck operators to connect their vehicles while stopped and use off-board power sources in order to operate on-board systems such as air conditioning, heating, and appliances, without any engine idling.”
Environmental concerns about diesel exhaust together with skyrocketing costs for diesel fuel are creating a very strong market for this type of service. This new article creates the necessary standardization of approach because a truck moves from jurisdiction to jurisdiction and needs to be able to connect to this infrastructure in any state in order for this to work.
626.10. Branch Circuits. Each stand must be supplied from a 208Y/120-V system or a 480Y/277-V system, with an exception for existing 120-V facilities
626.11. Feeder and Service Load Calculations. Each parking space must be calculated on the basis of not less than 11 kVA each, although Part (B) applies a demand factor to this load based on expected heating and air-conditioning burden. This is related to the “USDA Plant Hardiness Zone Map” and decreases from a high of 70 percent (Fairbanks Alaska) to a low of 20 percent (Houston, Texas) with some small increases to 24 percent for the highest zones (Miami, Florida and Honolulu, Hawaii) where an increased air conditioning load would take over from decreased heating load. The selection of a plant-hardiness map based on worst-case winter temperature is probably appropriate in terms of predicting worst-case winter loading, but may prove inaccurate in terms of predicting summer air conditioning load. It would seem to have been questionable in terms of predicting maximum demand, but data submitted with the proposal that tabulated actual measured demand from existing facilities showed close agreement with table predictions. These facilities are typically 100 percent occupied at night and therefore all available spaces go into the calculation prior to applying the demand factor.
Note that the 11 kVA is figured on the basis of the maximum power capability of the receptacles required by 626.24(B), that is, two 20-A receptacles on 120-V circuits, and one 30-A circuit on a 208-V circuit, as follows:
2(120 V × 20 A) + (208 V × 30 A) = 11,000 VA
Note that some gantry operations with umbilical assemblies drop from overhead and supply recirculated air from the truck cab after heating or cooling it as required. In practice such heating and air conditioning units in the gantry are not centralized, but remain on a site-by-site basis, because each truck is charged on the basis of specific services provided by the site operator. In these cases the heating and air conditioning is still part of the electrical load for each site, although it will not appear in the truck cab as a receptacle. As long as the feeder to a group of sites includes within its load profile the heating and air conditioning load for each of those sites, this calculation and the demand factors that follow will be correct.
626.22. Wiring Methods and Materials. If the supply is from a pedestal or raised concrete pad, the mounting height must be at least 600 mm (2 ft) above grade or above the prevailing high-water mark in areas subject to flooding. Supplies that drop from a gantry are obviously not subject to this limitation. The supply equipment must be accessible through an entryway not smaller than 600 mm (2 ft) wide and 2 m (6½ ft high) There must be a remote, permanently lock-open capable disconnecting means, readily accessible, that will open the supply to one or more spaces.
626.24. Electrified Truck Parking Space Supply Equipment Connection Means. This section begins with a requirement for extra-hard service cords to each connection, run together as a “single separable power supply cable assembly.” The receptacle requirements are unusual, and in at least some cases probably not written correctly. For example, consider the requirement in (B)(1) for two single receptacles, each two-pole three-wire grounding and connected to an individual branch circuit. Note that (D) requires that any receptacle outlets installed in accordance with this section have GFCI protection. No single GFCI receptacles are now in production, so this would mandate protection in the form of a GFCI circuit breaker at a considerable distance away, wherever this distribution originates.
There is a widely circulated set of full-color photographs of one of these umbilical-supplied assemblies that drop from an overhead gantry and are designed to mount in the passenger-side window. They are equipped with a conditioned air supply, a touch-screen computer terminal with a USB port and a card stripe reader, Ethernet port, etc. It also has two receptacles, at least of the kind the manufacturer apparently thought were two receptacles, because actually there are two GFCI duplex receptacles, for a total of four receptacles in violation of (1). But wait. If you look at the outside photo carefully, there is a double flap wet location cover for both halves of a duplex receptacle, not GFCI, with no provisions for it to be weatherproof while in use. This is where you plug in the block heater, as it turns out. It is apparent that we have quite a way to go before the equipment on the ground meets the NEC, and frankly, to where the NEC should be in terms of fine tuning this article.
This section also recognizes a single receptacle, “3-pole 4-wire grounding-type, single-phase rated either 208Y/120 V or 125/250 V.” Since there is no NEMA configuration for a 208Y/120 single phase receptacle at any amperage, this is confusing at best. There is a fine print note following that describes a standard for pin-and-sleeve devices. However, that is not a requirement, and 30 ampere 125/250-V plugs and receptacles are used by the million on identical distributions as these. The majority of multifamily housing is supplied through 208Y/120-V three-phase services, and almost without exception, the feeder to each apartment consists of two-phase conductors and a neutral. Every conventional dryer receptacle outlet will have one of these devices providing the same sort of connectivity on the identical distribution system. It appears the reference to 208 V may have been an attempt by a proprietary interest to game the process and should be ignored. At the inception of the 2008 NEC there is very little in the way of an installed user base and so there is no consensus as to what these receptacles should be.
This receptacle is intended to supply heating and air-conditioning equipment in the truck cab. If the conditioned air will be supplied by the truck stop through an umbilical connection instead, then this receptacle need not be provided.
626.25. Separable Power-Supply Cable Assembly. These are the rules, not inconsistent with the receptacle configuration rules in 626.24, that govern the power supply cords run from the site outlets to the truck. 15-A assemblies are permitted to operate an engine block heater on existing vehicles. The overall length of the cord is to be 7.5 m (25 ft) unless, if longer, a listed cable management system is employed.
626.26. Loss of Primary Power. This section and the next directly correspond with 625.25 and 625.26. It is not clear why these sections, which cover battery-powered vehicles for which there could be a use for the energy stored in the battery to power dwelling unit loads either on a stand-alone basis (625.25) or possibly in parallel with the utility (625.26), are being duplicated at a diesel truck stop. However, they do no harm.
626.30. Transport Refrigerated Units. These are the refrigerated trailers that are one of the principal reasons for diesel trucks to sit idle, and they represent a very significant load. This load is not included in the site load calculation of 626.11. There are two voltage options given here, either 30-A on a 480-V 3-phase system or 60-A on a 208-V 3-phase system, so these are significant power loads. The cord connections will be through extra-hard usage assemblies with a 90°C conductor rating along with an outer jacket evaluated for sunlight resistance and wet locations, and additional ratings for cold weather, oil and gasoline, ozone, acids, other chemicals, and abrasion.
630.1. Scope. There are two general types of electric welding: arc welding and resistance welding. In arc welding, an arc is drawn between the metal parts to be joined together and a metal electrode (a wire or rod), and metal from the electrode is deposited on the joint. In resistance welding, the metal parts to be joined are pressed tightly together between the two electrodes, and a heavy current is passed through the electrodes and the plates or other parts to be welded. The electrodes make contact on a small area—thus the current passes through a small cross section of metal having a high resistance—and sufficient heat is generated to raise the parts to be welded to a welding temperature.
In arc welding with AC, an individual transformer is used for each operator; in other words, a transformer supplies current for one arc only. When DC is used, there is usually an individual generator for each operator, though there are also multioperator arc-welding generators. Note that the scope also includes plasma cutting operations where the electrical equipment is involved in creating and maintaining the arc that creates the ionized gas that does the cutting.
630.11. Ampacity of Supply Conductors. The term transformer arc welder is commonly used in the trade and, hence, is used in the Code, though the equipment might more properly be described as an arc-welding transformer. Refer to the FPN following 630.31, where the term duty cycle is explained.
It is evident that the load on each transformer is intermittent. Where several transformers are supplied by one feeder, the intermittent loading will cause much less heating of the feeder conductors than would result from a continuous load equal to the sum of the full-load current ratings of all the transformers. The ampacity of the feeder conductors may therefore be reduced if the feeder supplies three or more transformers. Note that if the value “ I 1eff” is provided on the welder name plate, then this value must be used instead of the value determined from the table. This value can also be calculated using the formula in the fine print note under 630.12(B). This is the effective current for the welder; it contrasts with “ I 1max” which is essentially the rated primary current, as further described in the same note.
630.12. Overcurrent Protection. Arc-welding transformers are so designed that as the secondary current increases, the secondary voltage decreases. This characteristic of the transformer greatly reduces the fluctuation of the load on the transformer as the length of the arc, and consequently the secondary current, is varied by the operator.
The rating or setting of the overcurrent devices specified in this section provides short-circuit protection. It has been stated that with the electrode “frozen” to the work, the primary current will in most cases rise to about 170 percent of the current rating of the transformer. This condition represents the heaviest overload that can occur, and, of course, this condition would never be allowed to continue for more than a very short time. However, rating of the OC protection can be based on 200 percent of the maximum value of supply current or primary current of the welder, at the maximum rated output (next higher standard size permitted). The OC device may be located at the welder or at the line-end of the supply circuit.
630.31. Ampacity of Supply Conductors. Subparagraph (A)(1) applies where a resistance welder is intended for a variety of different operations, such as for welding plates of different thicknesses or for welding different metals. In this case, the branch-circuit conductors must have an ampacity sufficient for the heaviest demand that may be made on them. Because the loading is intermittent, the ampacity need not be as high as the rated primary current. A value of 70 percent is specified for any type of welding machine that is fed automatically. For a manually operated welder, the duty cycle will always be lower and a conductor ampacity of 50 percent of the rated primary current is considered sufficient.
example 1 A spot welder supplied by a 60-Hz system makes 400 welds per hour, and in making each weld, current flows during 15 cycles.
The number of cycles per hour is 60 × 60 × 60 = 216,000 cycles.
During 1 h, the time during which the welder is loaded, measured in cycles, is 400 × 15 = 6000 cycles.
The duty cycle is therefore (6000/216,000) × 100 = 2.8 percent.
example 2 A seam welder operates 2 cycles “on” and 2 cycles “off,” or in every 4 cycles the welder is loaded during 2 cycles.
The duty cycle is therefore 2/4 × 100 = 50 percent.
Transformers for resistance welders are commonly provided with taps by means of which the secondary voltage, and consequently the secondary current, can be adjusted. The rated primary current is the current in the primary when the taps are adjusted for maximum secondary current.
When a resistance welder is set up for a specific operation, the transformer taps are adjusted to provide the exact heat desired for the weld; then in order to apply subparagraph (A)(2), the actual primary current must be measured. A special type of ammeter is required for this measurement because the current impulses are of very short duration, often a small fraction of a second. The duty cycle is controlled by the adjustment of the controller for the welder.
The procedure in determining conductor sizes for an installation consisting of a feeder and two or more branch circuits to supply resistance welders is first to compute the required ampacity for each branch circuit. Then the required feeder ampacity is 100 percent of the highest ampacity required for any one of the branch circuits, plus 60 percent of the sum of the ampacities of all the other branch circuits.
Some resistance welders are rated as high as 1000 kVA and may momentarily draw loads of 2000 kVA or even more. Voltage drop must be held within rather close limits to ensure satisfactory operation.
630.32. Overcurrent Protection. In this case, as in the case of the overcurrent protection of arc-welding transformers (630.12), the conductors are protected against short circuits. The conductors of motor branch circuits are protected against short circuits by the branch-circuit overcurrent devices and depend on the motor-running protective devices for overload protection. Although the resistance welder is not equipped with any device similar to the motor-running protective device, satisfactory operation of the welder is a safeguard against overloading of the conductors. Overheating of the circuit could result only from operating the welder that either the welds would be imperfect or parts of the control equipment would be damaged, or both.
630.42. Installation. This section governs the placement of welding cables in cable tray, which must then be labeled accordingly. This cable is not fine-stranded building wire that happens to be easy to bend and install in motor control centers, etc. The UL data on this cable clearly limits its use to the “secondary circuit of electric welders” and is as follows:
This category covers welding cable, which is a single-conductor cable intended for use in the secondary circuit of electric welders in accordance with Art. 630, Part IV of ANSI/NFPA 70, “National Electrical Code.” The conductors are flexible-stranded copper, 8 AWG through 250 kcmil, the individual strands of which are 34 through 30 AWG.
Welding cable is rated 60, 75 or 90°C and 100 or 600 V.
The voltage and temperature ratings, if higher than 100 V and 60°C, respectively, are identified by printing on the surface of the insulation.
640.1. Scope. Centralized distribution systems consist of one or more disc or tape recorders and/or radio receivers, the audio-frequency output of which is distributed to a number of reproducers or loudspeakers.
A public-address system includes one or more microphones, an amplifier, and any desired number of reproducers or speakers. A common use of such a system is to render the voice of a speaker clearly audible in all parts of a large assembly room.
640.2. Definitions. “Abandoned Audio Distribution Cable” is previously installed cable that is not terminated at equipment and not identified for future use with a tag.
“Audio Signal Processing Equipment” essentially covers the range of equipment covered by this article, and the note that follows gives a full picture of the coverage, Note that a sound signal processor such as a computer creating sound signals from a MIDI (musical instrument digital interface) file is included.
“Technical Power System” covers systems using isolated grounding as covered in 250.146(D). However, the terminology is intriguing because it appears in another location in the NEC, namely, 647.6(B) and 647.7(A)(2). The provisions of Art. 647 are predominantly the creation of an audio engineer who was looking for a way to eliminate audio hum in recording studios. He succeeded so well that the grounding system in that article (center-tapped 120 V operating with both circuit wires running 60 V to ground), which started out in Art. 530, is now a stand-alone article with provisions to balance loads across a three-phase distribution. It can certainly be used to power Art. 640 applications, and with far more effective results than simply relying on 250.146(D), whose effectiveness is now widely questioned.
640.3. Locations and Other Articles. In general, the power-supply wiring from the building light or power service to the special equipment named in 640.1, and between any parts of this equipment, should be installed as required for light and power systems of the same voltage. Certain variations from the standard requirements are permitted by the following sections. For radio and television receiving equipment, the requirements of Art. 810 apply except as otherwise permitted here.
640.9 covers wiring to loudspeakers and microphones and signal wires between equipment components—tape recorder or record player to amplifier, and so on. As shown in Fig. 640-1, amplifier output wiring to loudspeakers handles energy limited by the power (wattage) of the amplifier and must conform to the rules of Art. 725. As shown in 725.41(A), the voltage and current rating of a signal circuit will establish it as either Class 2 or Class 3 signal circuit. Amplifier output circuits rated not over 70 V, with open-circuit voltage not over 100 V, may use Class 3 wiring as set forth in 725.41(A).
Fig. 640-1. Sound-system speaker wiring may be either Class 2 or 3 signal system, but must not intermingle with those systems. (Sec. 640.9.)
Article 725 of the Code covers, among other things, signal circuits. A signal circuit is defined as any electrical circuit which supplies energy to a device—such as a loudspeaker or an amplifier—that gives a recognizable signal.
640.6. Mechanical Execution of Work. Part (A) is the usual limited energy article requirement requiring neat and workmanlike work. Part (B) requires cables that are exposed (including above a hung ceiling) to be supported such that normal building use will not damage them. Above the hung ceiling, the cabling must meet 300.11(A), which means staying off ceiling support wires that are an integral support of the ceiling system, but additional wires are allowed; refer to that discussion in Chap. 3 for more information. Abandoned cables (not identified for future use with a tag) must be removed. Tagged cables left in place must have tags that will hold up in their environment and the tags must give the date of identification, the date of intended use, and information relative to the intended use.
640.7. Grounding. Part (A) has the usual grounding requirements for wireways and auxiliary gutters. Parts (B) and (C) address implementation of Art. 647. See the comments offered here in the definition of “Technical Power System” for more information.
640.8. Grouping of Conductors. In this class of work, the wires of different systems are in many cases closely associated in the apparatus itself; therefore, little could be gained by separating them elsewhere.
640.9. Wiring Methods. Part (A)(1) requires wiring connected to the premises wiring system to comply with Chaps. 1 through 4 except as modified.
Part (A)(2) requires separately derived systems to comply with Code rules generally. This subsection also recognizes the procedures in Art. 647, which effectively allows the full 60/120-V system procedures to be used in the context of this article.
Part (A)(3) requires all other wiring follow the rules in Art. 725.
Part (B) requires auxiliary power supply wiring to equipment with a separate input therefore must follow Art. 725 rules; batteries must be installed in accordance with Art. 480. “Auxiliary inputs mean that normal premises wiring supplies are also intended as a supply. An UPS is not an auxiliary input unless it is a direct part of the equipment and providing a dc supply.
Part (C) allows amplifier output wiring to follow the normal rules in Art. 725 for the Class of circuit generated, whether Class 1, 2, or 3, (and the amplifier must be listed and marked accordingly) except that common raceways or enclosures must contain only audio circuits of the same class (Fig. 640-1). Note that there are subtle differences in the capacities of audio power supplies and output circuits that, while allowing them to use limited energy wiring protocols, make them unsuitable for intermixing in common raceways. Therefore even in cases where you have a Class 2 audio output, that circuit may not be run in the same raceway as a normal Class 2 signaling circuit. This is addressed directly in 725.139(F).
Part (D) requires audio transformers and autotransformers to be used so as not to exceed the product limitations. System grounding isn’t required for circuits on the load side of this equipment, in contrast with the usual requirement for power circuits in 210.9.
640.10. Audio Systems near Bodies of Water. This section covers audio systems near bodies of water. An exception waives the requirements for watercraft, even if supplied by shore power. Wiring on these vessels would be exempt anyway, since Sec. 90-2(b)(1) provides that such wiring is beyond the scope of the Code. The rules in this section turn on whether the equipment is supplied by “branch-circuit power.” If so, it has to be kept at least 5 ft away from the water; otherwise, if rated as Class 2, then it is limited only by the manufacturer’s instructions.
640.21. Use of Flexible Cords and Cables. Cords used for connection to branch circuits are permitted according to the normal rules. Cords running between amplifiers and speakers, or speaker to speaker, or equipment to equipment, follow applicable Art. 725 requirements, with an allowance for other cabling.
In addition, for equipment to equipment applications, other cabling can be used if specified by the manufacturer. Battery and other power sources need to follow the applicable code rules. The note points out that some of these additional sources will end up being the only source, but that they may be supplied in turn by intermittent or continuous power from a branch circuit.
Part (E) permits cords to be used to connect permanently installed equipment racks to facilitate equipment access or for “isolating the technical power system of the rack from the premises ground.” This relates to the definition of “Technical Power System” and therefore appears intended to refer to isolation in the sense of connecting to an isolated-ground receptacle, as covered in 250.146(D). That wiring protocol still requires an equipment grounding conductor running with the circuit conductors. Therefore this would refer to a grounded cord plugged into an isolated ground receptacle. That receptacle might be supplied, however, by a “Technical Power System.” Refer to the discussion at 640.22, below.
640.22. Wiring of Equipment Racks and Enclosures. This covers the wiring of equipment racks, which if made of metal (the usual condition) must be grounded. The racks have to be neat and workmanlike and there needs to be “reasonable access” to equipment power switches and overcurrent protective devices. The supply cords have to terminate within the equipment rack enclosure in an “identified connector assembly.” They have to be able to carry the load and have overcurrent protection, which could be on the branch circuit.
Bonding isn’t required if the rack is supplied by a “technical power ground.” (Fig. 640-2) Sec. 640-2 defines a “Technical Power System” as one in which insulated grounding receptacles are used in accordance with 250.146(D). However, 640.7(B) specifically refers to 647.6, which mandates the use of the term “Technical Equipment Ground” in reference to separately derived 120-V systems operating with a midpoint ground (and therefore 60/120 V). Since the original proposal to add this system to the NEC came from the audio engineering community, probably the reference here to a “technical power ground” is indeed to a system covered by Art. 647.
Fig. 640-2. Grounding in racks may correlate with Art. 647 provisions. (Sec. 640.22.)
640.24. Wireways and Auxiliary Gutters. Wireways and auxiliary gutters follow the normal fill requirements; the 75 percent allowance in all versions of this article prior to the 1999 rewrite was not reinstated.
640.25. Loudspeaker Installation in Fire Resistance-Rated Partitions, Walls, and Ceilings. Speakers placed in fire-resistant partitions must be listed for the purpose or else go into enclosures (or a recess in the wall) that will maintain the fire-resistance rating.
640.41. Multipole Branch-Circuit Cable Connectors. Multipole branch-circuit cable connectors must be polarized and must not transmit strain to terminations, and the female half must be attached to the load end of the power supply cord. Differently rated devices must be uniquely configured so they can’t inter-mate, and they must not intermate with nonlocking devices used for speaker connections, or with connectors rated 250 V whether locking or not. This suggests 125-V locking configured devices could be acceptable for speaker connections.
Signal cabling not intended for such speaker connections must not be inter-mateable with multipole branch-circuit cable connectors of “any accepted configuration.” Note that this first part of this section substantially duplicates the language found in 520.67.
640.42. Use of Flexible Cords and Cables. This section covers the use of flexible cord for portable and temporary audio system applications, and Parts (A) through (D) are essentially the same as Sec. 640-21 for permanent installations. Part (E) requires that flexible cord run to supply a portable equipment rack must use listed extra-hard usage cord. For outdoor use, the cord must also be suitable for wet locations and sunlight resistant. If the racks also supply lighting or power equipment, Arts. 520 and 525 apply as appropriate. The use of any cable extensions, adapters, and breakout assemblies must meet the applicable provisions of those articles as well, because of the references in 640.3(F) and (G). Note that the relevant requirements principally appear in 520.68(A)(4) and 520.69.
640.43. Wiring of Equipment Racks. This section covers wiring of portable or temporary equipment racks, and it is similar to Sec. 640-22 which covers the permanent variety. There are significant differences, however. If the rack is nonmetallic and equipped with a cover, removal of that cover must not allow access to Class 1, Class 3, or “primary circuit power” without removal of terminal covers or the use of tools. Wiring that leaves such equipment to other equipment or a power supply must have strain relief or other arrangements so pull on the cord won’t increase the risk of damage.
640.44. Environmental Protection of Equipment. This section requires protection of portable or temporary equipment used outdoors from adverse weather conditions. If the equipment is expected to remain operational, arrangements need to be made including ventilating heat producing equipment.
640.45. Protection of Wiring. If there is public accessibility, cord can’t be laid on the ground or floor without being covered with approved nonconductive mats and there must not be a tripping hazard. This is essentially the same as 525.20(G), although the waiver of 300.5 allowing for temporary shallow burial is missing.
640.46. Equipment Access. Equipment likely to be hazardous to the public needs to be protected with barriers or supervised by qualified personnel to prevent public access.
645.1. Scope. This article applies only to an information technology equipment room. Because this article covers all of the designated “equipment,” “wiring,” and “grounding” that is contained in a room, there is no question about the mandatory application of these rules to such “rooms.” But the specific nature of that word room in the 645.1 statement of “Scope” leaves some uncertainty about electronic computer/data processing equipment and systems that are not installed in a dedicated room. In such cases, one must follow all the rules given in Chaps. 1 through 4 and ignore the rules and permissions given in Art. 645. The point being: If you don’t have a computer room, as defined by 645.4, then Art. 645 does not cover the installation, because, as given in the first sentence of 645.1, only equipment and the like installed in “an information technology equipment room” is covered.
Furthermore, this article is unusual in that it is, in effect, voluntary. If you do not want to comply with its provisions, simply make sure that one (or more) of the conditions in 645.4 is not met, and the article does not and never will apply. The article offers certain wiring advantages that may be beneficial, but there is no obligation to implement those advantages.
645.4. Special Requirements for Information Technology Equipment Room. Five conditions are described that must be complied with in order for Art. 645 to be applied to data processing equipment rooms:
1. One or more “grouped and identified” disconnects must be provided to open the supply to all electronic equipment and all HVAC equipment in the computer room, with this disconnect (or these disconnects) controlled at the “principal exit doors” of the room.
2. A dedicated HVAC system must be used for the room, or strictly limited use involving automatic actuation of fire/smoke dampers may be made of an HVAC system that “serves other occupancies.”
3. Only “listed” (such as by UL) information technology equipment may be installed.
4. The computer room must be “occupied only by those personnel needed for operating and maintaining the computer/data processing equipment.”
5. The room must have complete fire-rated separation from “other occupancies.”
It is very clear from the wording that if any of these conditions is not met, the entire computer/data processing installation is not subject to the rules of Art. 645. But such an installation would be subject to all other applicable rules of the NEC.
645.5. Supply Circuits and Interconnecting Cables. Part (A) limits every branch circuit supplying data processing units to a maximum load of not over 80 percent of the conductor ampacity (which is an ampacity of 1.25 times the total connected load).
Part (B) covers use of computer or data processing cables and flexible cords, which can be up to 4.5 m (15 ft) long, and must be protected from damage if run across the floor. As shown in Fig. 645-1 (under a raised floor), part (C) permits data processing units to be “interconnected” by flexible connections that are “listed for the purpose.”
Fig. 645-1. Connection of data-processing units to their supply circuits and interconnection between units (power supply, memory storage, etc.) may be made only with cables or cord-sets specifically approved as parts of the data processing system. (Sec. 645.5.)
Part (D) permits a variety of wiring methods under an accessible [condition (1)] raised floor serving a data processing system: metal surface raceway with metal cover, metal wireway (as shown in Fig. 645-2), liquidtight flexible conduit, rigid metal conduit, EMT, flexible metal conduit, IMC, Type MI cable, Type AC cable (commonly known as “BX”), and Type MC cable (Fig. 645-2), as well as many others, both metallic and nonmetallic [condition (2)]. Note also that Tray Cable, Type TC, is also here through the back door through 645.5(D)(6)(c) because it is in the table. This not only allows various wiring methods, it also allows receptacles associated with the data processing equipment under the floor. However, until a 2008 NEC change, technically you couldn’t run a cord through the floor to plug into a receptacle that was allowed below the floor. This has been fixed, although the wording syntax needs improvement. Item (3) mentions supply cords per 645.5(B), but not in the form of a statement of a condition to be met, which the parent language in 645.5(D) requires. This will need to be corrected in the next code cycle.
Fig. 645-2. Branch circuits from a panel-board to data processing receptacle outlets can be in a wide variety of Chap. 3 wiring methods. (Sec. 645.5.)
Although 386.12 prohibits use of a metal surface raceway where it would be concealed, part (3) under 386.10 recognizes its use under raised flooring for data processing by referencing Art. 645. It should be noted that, in addition to a surface metal raceway, a metal wireway may be used under a raised floor. 645.5(D)(2) does recognize use of “wireway” under raised floors. 376.10(1) accepts wireway “for exposed work,” but it may be used under raised floors and above suspended ceilings of lift-out tiles because of the definition of “exposed.” In addition, 300.22(C) does recognize metal wireway above a suspended ceiling space used for air-handling purposes.
It’s worth noting that while the NEC may permit the use of nonmetallic race-ways and wireways under the raised floor in an information technology room, local building codes may prohibit such installation. Check with the local building department or building inspector for guidance on any prohibition on the use of nonmetallic or nonmetallic-coated raceways, enclosures, and wire-ways. As covered in the next paragraph, those prohibitions are now obsolete and should not be continued.
This building code problem became critical when some cables used under raised floors were rejected by building inspectors as not being of plenum grade, and they were treating the floor as a plenum cavity. However, the reason these cables were used in the first place was that they had certain electrical characteristics in terms of limited capacitive reactance that made them essential in comparison to plenum grade cabling. Wiring methods could be changed, but this meant that the equipment would not even function properly. This direct conflict between construction disciplines set the stage for an extended standoff that continued for years. The solution was arrived at through extended discussion between electrical code experts and leaders of the model building code organizations, and it is reflected in 645.5 (D)(4). The building officials decided that they would not need to classify the underfloor space as a plenum cavity if it would not function in that way during a fire condition. The electrical experts said, in effect, done! There was already a drop-out relay under these floors to shut down the air circulation, as required by 645.10. All that was required was to add a smoke detector under the floor and have it drop out the ventilation if products of combustion were detected. The fan would stop, the air movement would cease, and the plenum cavity designation need not apply.
Part (D)(5) requires openings in raised floors to provide abrasion protection for cables passing up through the floor and requires that openings be only as large as needed and made in such a way as to “minimize the entrance of debris beneath the floor.” Part (D)(6) covers direct cabling, especially limited energy cabling as covered in the new Table 645.5. These cables, along with others listed as “Type DP” are acceptable for these locations.
Part (E) adds important information by stating clearly that any cable, boxes, connectors, receptacles, or other components that are “listed as part of, or for, information technology equipment” are not required to be secured in place, but any cable or equipment that is not “listed as part of” the computer equipment must be secured in accordance with all Code rules covering them.
Part (F) requires that the accessible portions of abandoned supply circuits and interconnecting cables, defined as unterminated and unidentified for future use, be removed. The next part (G) builds on this by requiring that in order to qualify as identified for future use, the cables needed to be tagged in a manner with sufficient durability for the location and stating the date of identification, the expected date of use in the future, and some information as to what that use is expected to be.
645.10. Disconnecting Means. As shown in Fig. 645-3, a master means of disconnect (which could be one or more switches or breakers) must provide disconnect for all computer equipment, ventilation, and air-conditioning (A/C) in the data processing (DP) room.
The disconnects called for in this rule are required to shut down the DP system and its dedicated HVAC and to close all required fire/smoke dampers under emergency conditions, such as fire in the equipment or in the room. For that reason, the rule further requires that the disconnect for the electronic equipment and “a similar” disconnect for A/C (which could be the same control switch or a separate one) must be grouped and identified and must be “controlled” from locations that are readily accessible to the computer operator(s) or DP manager. And then the rule specifies that these one or more emergency disconnects must be installed at all “principal” exit doors—any doors that occupants of the room might use when leaving the DP room under emergency conditions. The concept is that operators would find it easy to operate the one or more control switches as they exited the room through the doorway. Figure 645-4 shows two control switches—one in the control circuit of the A/C system and the other a shunt-trip pushbutton in the CB of the feeder to the DP branch circuits—with a collar guard to prevent unintentional operation.
Fig. 645-3. Data processing room must have arrangements like those shown here. (Sec. 645.10.)
Although the present wording of this rule readily accepts the use of a single disconnect device (pushbutton) that will actuate one or more magnetic contactors that switch the feeder or feeders supplying the branch circuits for the computer equipment and the circuits to the A/C equipment, the wording also recognizes the use of separate disconnect control switches for electronic equipment and A/C. Control of the branch circuits to electronic equipment may be provided by a contactor in the feeder to the transformer primary of a computer power center, as shown in Fig. 645-5.
Fig. 645-4. Adjacent to the door of a DP room, a break-glass station (at top) provides emergency cutoff of the A/C system in the DP room; and a mushroom-head pushbutton—with an extended collar guard that requires definite, intentional pushing action—energizes a shunt-trip coil in the feeder circuit breaker supplying branch circuits for the electronic DP equipment. (Sec. 645.10.)
Fig. 645-5. Rapidly accelerating application of DP equipment in special DP rooms with wiring under a raised floor of structural tiles places great emphasis on Art. 645. “Computer power centers” (arrow) are complete assemblies for the supply of branch circuits to DP equipment, with control, monitoring, and alarm functions. (Sec. 645.10.)
A single means used to control the disconnecting means for both the electronic equipment and the air-conditioning system offers maximum safety. In the event of a fire emergency, having two separate disconnecting means (or their remote operators) at the principal exit doors will require the operator to act twice and thus increase the hazard that only the electronic equipment or the air-conditioning system will be shut down. If only the electronic equipment is disconnected, a smoldering fire will become intensified by the air-conditioning system force-ventilating the origin of combustion. Similarly, if only the air-conditioning system is disconnected, either a fire within the electronic equipment will become intensified (since the electric energy source is still present) or the electronic equipment could become dangerously overheated due to the lack of air conditioning in this area.
Wording of this rule requires means to disconnect the “dedicated” A/C “system serving the room.” If the DP room has A/C from the ceiling for personnel comfort and A/C through the raised floor space for cooling of the DP equipment, both A/C systems would have to be disconnected. There have been rulings that only the A/C serving the raised floor space must be shut down to minimize fire spread within the DP equipment and that the general room A/C, which is tied into the whole building A/C system, does not have to be interrupted. In other cases, it has been ruled that the general room A/C must be shut down, while the floor space A/C may be left operating to facilitate the dispersion of fire suppressant and extinguishing materials within the enclosures of DP equipment that is on fire. Because of the possibility of various specific interpretations of very general rules, this whole matter has become extremely controversial. Review, as well, the history behind the present wording on 645.5(D)(4) which bears on this question.
The exception to this rule waives the need for disconnect means in any “integrated electrical systems” (Art. 685), where orderly shutdown is necessary to ensure safety to personnel and property. In such cases, the entire matter of type of disconnects, their layout, and their operation is left to the designer of the specific installation.
645.11. Uninterruptible Power Supplies (UPSs). This rule requires disconnects for “supply and output circuits” of any UPS “within” the computer room. The UPS disconnecting means must satisfy 645.10, and it must “disconnect the battery from its load.” The wording of this rule leaves questions about a disconnecting means for a UPS installed outside the computer room (Fig. 645-6). The UPS output circuit—which presumably feeds the computer room—would have to be disconnected. But it seems as if the supply to the UPS and the battery charge/supply circuit need not be automatically disconnected when the UPS is located outside the room. On the other hand, the wording clearly requires the UPS output circuit to meet 645.10 even if the battery was not required to be disconnected from its load because the unit, in this case, is outside the room.
645.15. Grounding. Data processing equipment must either be grounded in full compliance with Art. 250 or be “double insulated.” However, immediately following this statement is an allowance that is so important it frequently drives the decision about whether to qualify for Art. 645 coverage or not. Power systems derived within listed IT equipment are not required to comply with 250.20(D), which sets in motion a chain of events that includes the permission to not bother installing a grounding electrode or making a connection to an existing electrode. The note following explains that listing requirements adequately address the safety issues involved. The principal issue is whether or not the power system includes a system bonding jumper between the neutral and the equipment grounding system between the output terminals of the transformer and the secondary disconnect. If this jumper is properly installed, then any electrical fault will have an effective fault current path to return to the system source. The product standards do ensure that this jumper will be installed.
Fig. 645-6. Considerable interpretation latitude is inherent in the rule that requires “grouped and identified” switches or circuit breakers “at principal exit doors” to control an uninterruptible power supply. (Sec. 645.11.)
The last sentence of 680.15 points out the need for bonding of “signal reference structures” where such grids are installed within the information technology room. Presumably this bonding should be provided between the equipment ground bus at the power systems source and the reference grid. As for sizing this bonding jumper, in the absence of a specific rule on this matter, it seems reasonable to size the bonding jumper based on the size of the source’s main or system bonding jumper. Where the room is supplied by a feeder, a bonding jumper not less than the size of the equipment grounding conductor required by 250.122 for the feeder would also seem to satisfy this rule. Such sizing should be viewed as acceptable especially since there is no wording provided to establish a method for the sizing of this required bonding conductor.
645.17. Power Distribution Units. This section recognizes the use of power distribution units (PDUs) that are equipped with multiple panelboards. That is, PDUs may have more than one panelboard within a single enclosure, provided each panelboard does not have more than 42 OC devices—as used to be generally prohibited by 408.35 for panelboards. Such PDUs must be listed as “utilization equipment” for “information technology application” to fully meet this requirement. That limitation has been removed for the 2008 NEC, so this limitation should be removed as well. As this is written, there is a “Tentative Interim Amendment” circulating at NFPA that would do just that, assuming it gets the required vote in the code making panel.
647.1. Scope. This article covers a new 60/120-V distribution system designed to eliminate the effects of electronic noise, originally on audio and video production when this material was a Part G in Art. 530, and now in a more broadly applicable format. This system, as shown in Fig. 647-1, uses two ungrounded circuit conductors at a potential of 120 V, and the system secondary is midpoint tapped and grounded to hold the system to 60 V to ground.
Fig. 647-1. A panelboard properly wired to distribute power using circuit conductors with 120 V between them and with each wire running 60 V to ground. The system must not be designed to be used for line to neutral load (Sec. 647.1).
These systems have proven extremely effective at reducing audio hum and video interference, which is why this system started its days of NEC recognition in Art. 530. The noise from the filtering circuits tends to cancel on this type of balanced system and it routinely outperformed the most elaborate grounding arrangements on traditional distributions. All harmonic frequencies cancel. The transformers with the center-tapped 120-V secondaries are increasingly available as well. The location here as a standalone article means that other applications can use its benefits.
647.3. General. These systems can be used to reduce electronic noise in “sensitive electronic equipment locations” provided it is installed only in industrial or commercial occupancies, and under close supervision of qualified personnel.
647.4. Wiring Methods. Part (A) provides that standard panelboards, such as 120/240 single-phase panelboards are permitted, but the voltage system must be clearly marked on the panel face or door. Common-trip two-pole circuit breakers must be used for all circuits (or a two-pole fused switch), and they must be identified for use at the system voltage. In general, circuit breakers work OK as long their voltage rating isn’t exceeded, and that won’t be a problem on this system. As shown in the drawings, all loads supplied by these systems will use two ungrounded circuit conductors.
Part (B) covers junction box covers, which must be clearly marked to indicate the distribution panelboard and the system voltage.
Part (C) covers the identification rules for feeder and branch-circuit conductors, which must be identified as belonging to this kind of system at all splices and terminations by color, tagging, or equally effective means. The means chosen must be marked on all panelboards, and on the building disconnect. This is intended to reduce the possibility of confusion during future alterations.
Part (D) is one of very few instances in the NEC of a mandatory voltage drop requirement. The voltage drop on any branch circuit must not exceed 1.5 percent, and the total drop including the feeder contribution must not exceed 2.5 percent. These circuits are operating with only one-half the voltage to ground, which means that in any ground fault, only one-half the fault current would flow across the same fault. The intent of the rule is to maximize the available voltage to ground, which is already starting out at only one-half the normal amount, so overcurrent devices will trip open as quickly as possible in the event of a ground fault. For receptacles the limits are even tougher, with 1 percent on the branch circuit and 2 percent total from source to outlet if there is an intervening feeder.
647.5Three-Phase Systems. To equalize the load on a three-phase distribution, these systems can be arranged with three transformers connected on their primary side to a premises conventional wye distribution with one primary transformer circuit (see. Fig. 647-1) supplied by each of the three phases (Fig. 647-2). A maximum of two disconnects per winding are permitted, for a total of six. Note that with two disconnects per winding, the primary side of the transformers would be limited to 125 percent protection per 450.3(B). The more usual case would be three secondary disconnects, one per phase winding. Six-phase transformers operating 60 V to the star point are commercially available to support these wiring systems, with the secondary windings identified as in Fig. 647-2.
Fig. 647-2. Three systems derived from three transformers connected to different phase pairs result in three different technical power systems, each one displaced 120 electrical degrees from the next and capable of supplying two disconnects each (Sec. 647.5)
647.6. Grounding. These are separately derived systems, and as shown in the drawing, they must meet the requirements in 250.30. The neutral connection at the center tap of the transformer has no circuit load function; its sole function is to stabilize the voltage to ground and to provide an equipment grounding return path. It supplies, on the downstream side of the main bonding jumper for the system, equipment grounding terminal bars. They must be marked “Technical Equipment Ground.”
If there were a succession of panelboards and the system designer decided to isolate the return path from the enclosing raceways and panelboard enclosures for further noise reduction, this section allows the use of insulated grounding conductors and having the Technical Equipment Grounding busbars isolated from the enclosures. This is the same procedure allowed in 250.146(D) and 408.40. Note, however, that such raceways and enclosures must still be equipment grounding system. This can happen using any of the recognized equipment grounding conductors in 250.118
Conversely, the last sentence allows for a raceway grounding return path without any additional grounding conductors, provided that the impedance of the grounding return path over the raceway does not exceed the impedance of a separate grounding conductor installed to meet these minimum requirements. This is why the equipment grounding conductors in Fig. 647-1 have the notation “(if required).” Remember, if the circuit conductors have been increased to meet the voltage-drop requirement, then 250.122 requires a corresponding increase in the size of grounding conductors on the same circuit. In this case the raceway impedance must be compared with the lower impedance of the larger equipment grounding conductor that would otherwise have been substituted. The FPNs that follow reinforce this concept.
647.7. Receptacles. Where receptacles are used to connect equipment, all of the following conditions need to be met. The reason for these requirements is to make as certain as possible that the receptacles aren’t used for cord- and plug-connected loads designed for use on systems with only one ungrounded conductor. For example, if a floor lamp were connected, the screw shell would remain alive at 60 V to ground, even with the switch off, assuming a single-pole lampholder:
(1) All 15- and 20-A receptacles shall be GFCI protected. This item then adds an additional level of safety by requiring GFCI protection at or ahead of all receptacle outlets. Be careful if you are considering using a 2-pole GFCI for this purpose. Its electronic circuitry may not work on a system voltage of only 60 V to ground. The best choice is the so-called “master” GFCI units that are, in effect, feed through GFCI receptacles without the contact slots. They will work because they would see only their design voltage of 120 V. Remember, using an actual GFCI receptacle would violate the configuration rule in (4).
(2) All outlet strips, adapters, receptacle covers, and faceplates shall be marked: “WARNING—TECHNICAL POWER”; “Do not connect to lighting equipment”; “For electronic equipment use only”; “60/120 V 1 jac”; and “GFCI protected.” These markings are self-explanatory.
(3) A conventional 125-V 15-A or 20-A receptacle must be located within 6 ft of all permanently-installed 15-A or 20-A 60/120-V technical power system receptacles. This rule relieves the temptation to plug the proverbial floor lamp into the wrong receptacle, because an appropriate receptacle must be nearby.
(4) All 125-V receptacles used for 60/120-V technical power must be uniquely configured and identified for use with this class of system. These configurations have yet to be developed. When this system first entered the NEC (1996 edition), an exception followed, allowing a conventionally configured receptacle in “machine rooms, control rooms, equipment rooms, equipment racks, and similar locations that are restricted to use by qualified personnel.” This is essential, because no such configuration has been developed. Unfortunately, the exception was converted into conventional text in a manner that does not properly convey the intent, and needs to be rewritten. However, the only appropriate action is to read the last sentence as the exception it started out as, and is still intended to be.
647.8. Lighting Equipment. Luminaires and associated equipment operated on these systems must have a disconnect that opens both circuit conductors. They must be permanently installed and listed for operation on these systems. No screw shell for such equipment is permitted to be exposed. Since fluorescent ballasts are a renowned source of electronic noise on wiring systems, placing them on one of these systems and thereby reducing that noise component to zero would benefit the overall system. However, the listing requirement may be excessive and unrealistic in terms of persuading conventional luminaire manufacturers to have their products reexamined.
650.4. Source of Energy. As now specified here, electronic organs must be supplied from a transformer-type of rectifier that operates at not more than 30 VDC.
650.6. Conductors. In Part (A) 28 AWG wiring for electronic signal circuits and 26 AWG for electromagnetic valve supply are recognized minimums. These instruments operate with common return conductors that may see currents from multiple sources at the same time, resulting in a 14 AWG minimum in the return from the electromagnetic supply. In part (C), the wires of the cable are normally all of the same polarity; hence, they need not be heavily insulated from one another. The full voltage of the control system exists between the wires in the cable and the common return wire; therefore, the common wire must be reasonably well-insulated from the cable wires.
650.7. Installation of Conductors. A 30-V system involves very little fire hazard, and the cable may be run in any manner desired; but for protection against injury and convenience in making repairs, the cable should preferably be installed in a metal raceway.
650.8. Overcurrent Protection. Circuits must be arranged so the 26 and 28 AWG wiring is protected at not over 6 A, with other size wiring protected at its ampacity. The common return wires do not require protection.
660.1. Scope. This covers industrial or other nonmedical applications exclusively; health care provisions are covered in Art. 517, Part V. An x-ray tube of the hot-cathode type, as now commonly used, is a two-element vacuum tube in which a tungsten filament serves as the cathode. Current is supplied to the filament at low voltage. In most cases, unidirectional pulsating voltage is applied between the cathode and the anode. The applied voltage is measured or described in terms of the peak voltage, which may be anywhere within the range of 10,000 to 1 million V, or even more. The current flowing in the high-voltage circuit may be as low as 5 mA or may be as much as 1 A, depending on the desired intensity of radiation. The high voltage is obtained by means of a transformer, usually operating at 230-V primary, and usually is made unidirectional by means of two-element rectifying vacuum tubes, though in some cases an alternating current is applied to the x-ray tube. The x-rays are radiations of an extremely high frequency (or short wavelength), which are the strongest in a plane at right angles to the electron stream passing between the cathode and the anode in the tube.
660.4. Connection to Supply Circuit. Fixed and stationary equipment must be connected through a Chap. 3 wiring method, with an allowance for 30 A and lesser ratings to use hard-service cord and a plug. Portable and relocatable equipment can use cord-and-plug connections up to a 60 A rating.
660.5. Disconnecting Means. The minimum rating is 50 percent of the short-time current rating or 100 percent of the long-time rating, whichever is higher, with an allowance for cord-and-plug connections at 30 A or less and on a 120 V branch circuit.
660.6. Rating of Supply Conductors and Overcurrent Protection. Branch circuit wiring is sized the same as the disconnecting means. Feeders supplying multiple units are sized on the basis of 100 percent of the largest two units (determined by making the comparison required in the branch-circuit analysis), plus 20 percent of all others.
660.20. Fixed and Stationary Equipment. In radiography it is important that the exposure be accurately timed, and for this purpose a switch is used which can be set to open the circuit automatically in any desired time after the circuit has been closed.
660.24. Independent Control. If multiple x-ray units or other equipment are operated from a single circuit operating over 600 V, each such piece of equipment must have its own switch or equivalent. This device shall be arranged and installed such that its live parts will not be contacted by persons using the equipment or others.
660.35. General. A power transformer supplying electrical systems is usually supplied at a high primary voltage; hence, in case of a breakdown of the insulation on the primary winding, a large amount of energy can be delivered to the transformer. X-ray units are supplied at lesser voltages and can do without the protections normally required for transformers in Art. 450.
660.47. General. This section definitely requires that all new x-ray equipment be so constructed that all high-voltage parts, except leads to the x-ray tube, are in grounded metal enclosures. Conductors leading to the x-ray tube are shielded and heavily insulated.
665.1. Scope. Induction and dielectric heating are systems wherein a work-piece is heated by means of a rapidly alternating magnetic or electric field.
Dielectric heating In contrast, dielectric heating is used to heat materials that are nonconductors, such as wood, plastic, textiles, and rubber, for such purposes as drying, gluing, curing, and baking. It uses frequencies from 1 to 200 MHz, especially those from 1 to 50 MHz. Vacuum-tube generators are used exclusively to supply dielectric heating power, with outputs ranging from a few hundred watts to several hundred kilowatts.
Whereas induction heating uses a varying magnetic field, dielectric heating employs a varying electric field. This is done by placing the material to be heated between a pair of metal plates, called electrodes, in the output circuit of the generator. When high-frequency voltage is applied to the electrodes, a rapidly alternating electric field is set up between them, passing through the material to be heated. Because of the electrical charges within the molecules of this material, the field causes the molecules to vibrate in proportion to its frequency. This internal molecular action generates the heat used for dielectric heating.
Induction heating Induction heating is used to heat materials that are good electrical conductors, for such purposes as soldering, brazing, hardening, and annealing. Induction heating, in general, involves frequencies ranging from 3 to about 500 kHz, and power outputs from a few hundred watts to several thousand kilowatts. In general, motor-generator sets are used for frequencies up to about 30 kHz; spark-gap converters, from 20 to 400 kHz; and vacuum-tube generators, from 100 to 500 kHz. Isolated special jobs may use frequencies as high as 60 to 80 MHz. Motor-generator sets normally supply power for heating large masses for melting, forging, deep hardening, and the joining of heavy pieces, whereas spark-gap and vacuum-tube generators find their best applications in the joining of smaller pieces and shallow case hardening, with vacuum-tube generators also being used where special high heat concentrations are required.
To heat a workpiece by induction heating, it is placed in a work coil consisting of one or more turns, which is the output circuit of the generator (Fig. 665-1). The high-frequency current which flows through this coil sets up a rapidly alternating magnetic field within it. By inducing a voltage in the workpiece, this field causes a current flow in the piece to be heated. As the current flows through the resistance of the workpiece, it generates heat (I2R loss) in the piece itself. It is this heat that is utilized in induction heating.
Fig. 665-1. A generator circuit supplies the work coil of an induction heater. (Sec. 665.2.)
665.5. Output Circuit. The induction circuit can be isolated from ground, as is often done with induction furnaces that employ water cooling. This mitigates the energy available in a ground fault until the furnace can be shut down. Otherwise the ground fault could compromise the water jacket. Water hitting molten metal will result in a steam explosion ejecting molten metal from the furnace and endangering anyone nearby. If not isolated, the system must be arranged so there is no shock hazard of over 50 V to ground on any accessible surface, even if the coil and heated object touch.
665.7. Remote Control. In part (A), if interlocking were not provided, there would be a definite danger to an operator at the remote-control station. It might then be possible, if the operator had turned off the power and was doing some work in contact with a work coil, for someone else to apply power from another point, seriously injuring the operator.
665.10. Ampacity of Supply Conductors. Quite often where several pieces of equipment are operated in a single plant, it is possible to conserve on power-line requirements by taking into account the load or use factor of each piece of equipment. The time cycles of operation on various machines may be staggered to allow a minimum of current to be taken from the line. In such cases the Code requires sufficient capacity to carry all full-load currents from those machines which will operate simultaneously, plus the standby requirements of all other units.
665.12. Disconnecting Means. An in-sight or lock-open disconnect must be provided for each heating equipment, sized to the equipment nameplate.
665.22. Access to Internal Equipment. This section allows the manufacturer the option of using interlocked doors or detachable panels. Where panels are used and are not intended as normal access points, they should be fastened with bolts or screws of sufficient number to discourage removal. They should not be held in place with any type of speed fastener.
665.25. Dielectric Heating Applicator Shielding. This section is intended primarily to apply to dielectric heating installations where it is absolutely essential that the electrodes and associated tuning or matching devices are properly shielded.
When it is necessary to transmit the high-frequency output of a generator any distance to the work applicator, a radio-frequency line is generally used. This usually consists of a conductor totally enclosed in a grounded metal housing. This central conductor is commonly supported by insulators, mounted in the grounded housing, and periodically spaced along its length. Such a line, rectangular in cross section, may even be used to connect two induction generators to the load.
While contact with high-voltage radio frequencies may cause severe burns, contact with high-voltage DC could be fatal. Therefore, it is imperative that generator output (directly, capacitively, or inductively coupled) be effectively grounded with respect to DC so that, should generator failure place high-voltage DC in the tank oscillating circuit, there will still be no danger to the operator. This grounding is generally internal in vacuum-tube generators. In all types of induction generators, one side of the work coil should usually be externally grounded.
In general, all high-voltage connections to the primary of a current transformer should be enclosed. The primary concern is the operator’s safety. Examples would be interlocked cages around small dielectric electrodes and interlocking safety doors.
On induction heating jobs, it is not always practical to completely house the work coil and obtain efficient production operation. In these cases, precautions should be taken to minimize the chance of operator contact with the coil.
665.26. Grounding and Bonding
At radio frequencies, and especially at dielectric-heating frequencies (1 to 200 MHz), it is very possible for differences in radio-frequency potential to exist between the equipment proper and other surrounding metal objects or other units of the complete installation. These potentials exist because of stray currents flowing between units of the equipment or to ground. Bonding is therefore essential, and such bonding must take the form of very wide copper or aluminum straps between units and to other surrounding metal objects such as conveyors and presses. The most satisfactory bond is provided by placing all units of the equipment on a flooring or base consisting of copper or aluminum sheet, thoroughly joined where necessary by soldering, welding, or adequate bolting. Such bonding reduces the radio-frequency resistance and reactance between units to a minimum, and any stray circulating currents flowing through this bonding will not cause sufficient voltage drop to become dangerous.
Shielding at dielectric-heating frequencies is a necessity to provide operator protection from the high radio-frequency potentials involved, and also to prevent possible interference with radio communication systems. Shielding is accomplished by totally enclosing all work circuit components with copper sheet, copper screening, or aluminum sheet.
668.1. Scope. This article provides effective coverage of basic electrical safety in electrolytic cell rooms.
The presentation of these requirements was accompanied by a commentary from the technical subcommittee that developed them. Significant background information from that commentary is as follows:
In the operation and maintenance of electrolytic cell lines, however, workmen may be involved in situations requiring safeguards not provided by existing articles of the NEC. For example, it is sometimes found that in the matter of exposed conductors or surfaces it is the man or his workplace which has to be insulated rather than the conductor. Work practices and rules such as are included in IEEE Trial Use Standard pertinent to such specific situations have been developed which offer the same degree of safety provided by the traditional philosophy of the NEC.
As a corollary to this concept, overheating of conductors, overloading of motors, leakage currents and the like may be required in cell lines to maintain process safety and continuity.
Proposed Art. 668 introduces such concepts as these as have been proven in practice for electrolytic cell operation.
668.2. Definitions. The subcommittee noted:
An electrolytic cell line and its DC process power supply circuit, both within a cell line working zone, comprise a single functional unit and as such can be treated in an analogous fashion to any other individual machine supplied from a single source. Although such an installation may cover acres of floor space, may have a load current in excess of 400,000 amperes DC or a circuit voltage in excess of 1000 volts DC, it is operated as a single unit. At this point, the traditional NEC concepts of branch circuits, feeders, services, overload, grounding, disconnecting means are meaningless, even as such terms lose their identity on the load side of a large motor terminal fitting or on the load side of the terminals of a commercial refrigerator.
It is important to understand that the cell line process current passes through each cell in a series connection and that the load current in each cell is not capable of being subdivided in the same fashion as is required, for example, in the heating circuit of a resistance-type electric furnace by Sec. 424.72(A).
668.3. Other Articles. Electrical equipment and applications that are not within the space envelope of the “cell line working zone,” as dimensioned in 668.10, must comply with all the other regulations of the NEC covering such work.
668.11. Direct-Current Cell Line Process Power Supply. These conductors are not required to be grounded. If that power supply is operating over 50 V between conductors, then the metal enclosures must be grounded either through protective relaying or by using a copper bonding conductor not less than 2/0 AWG in size.
668.12. Cell Line Conductors. These can be of any suitable conductive material, joined by bolting, welding, clamping, or compression, and of such size that under maximum load and ambient temperature their insulating supports will not be unsafely heated.
668.13. Disconnecting Means. As shown in Fig. 668-1, each DC power supply to a single cell line must be capable of being disconnected. And the disconnecting means may be a removable link in the busbars of the cell line.
Fig. 668-1. Removal of busbar sections may provide disconnect of each supply. (Sec. 668.13.)
668.14. Shunting Means. Similar means as for disconnecting can be used to make a shunted current path around one or more cells.
668.15. Grounding. The requirements of Art. 250 apply to any equipment or structure or other apparatus required to be grounded, however, an otherwise qualified water pipe electrode is not required to be included in the grounding electrode system.
668.20. Portable Electrical Equipment. This section and the rules of 668.21, 668.30, 668.31, and 668.32 cover installation and operating requirements for cells with exposed live conductors or surfaces. These rules are necessary for the conditions as noted by the subcommittee:
In some electrolytic cell systems, the terminal voltage of the cell line process power supply can be appreciable. The voltage to ground of exposed live parts from one end of a cell line to the other is variable between the limits of the terminal voltage. Hence, operating and maintenance personnel and their tools are required to be insulated from ground. If the cell-line voltage does not exceed 200 V dc, grounded enclosures are permitted but not required, and such enclosures are permitted to be grounded where guarded. To this end, receptacle circuits that supply this equipment as used within the cell-line working zone must only use ungrounded conductors in their supply created by the use of isolating transformers with ungrounded secondaries. Such tools and equipment must be marked to this effect, and the plugs and receptacles must be so configured that they will not be connectable to conventional receptacles, nor inadvertently interchanged with other equipment designed for conventional connections.
668.21. Power-Supply Circuits and Receptacles for Portable Electrical Equipment. Part (A) reiterates the requirement to use isolating power supplies for the ungrounded receptacles in the cell line working zone, and adds the requirement that the transformer primary cannot operate over 600 V with appropriate overcurrent protection. The ungrounded secondary conductors must also have overcurrent protection, and not be operating over 300 V.
668.30. Fixed and Portable Electrical Equipment. AC systems in the cell line working zone, and all exposed conductive surfaces, are not required to be grounded. The wiring to such equipment must be by flexible cord or nonmetallic raceway or cable assemblies. If metal raceways of cables are used, insulating breaks must be installed to avoid a hazardous condition. Fixed electrical equipment operating in the cell line zone is permitted to be bonded to the energized conductive surfaces of the cell line, and required to be so if it is attached to such a surface. Control and instrumentation circuits operating within the cell line working zone does not require overcurrent protection.
668.32. Cranes and Hoists. Conductive surfaces that enter the cell line working zone are not required to be grounded, and any part that contacts an energized cell or equipment attached to such a cell must be insulated from ground. Remote crane or hoist controls must employ (1) isolated circuits; or (2) pendants with nonconductive supports and surfaces, or ungrounded surfaces; or (3) radio controls; or (4) a rope operator.
668.40. Enclosures. General-purpose enclosures are permitted where natural drafts prevent accumulations of gas from the process.
669.1. Scope. This article covers all electroplating activity including anodizing. These operations often involve very high current at relatively low voltage, to the point that conventional conductors are not practicable.
669.5. Branch-Circuit Conductors. This wiring must be rated for 125 percent of the connected load; busbar ampacity is that of busbars in auxiliary gutters in 366.23.
669.6. Wiring Methods. For systems not over 50 V dc, open wires are permitted without insulated support if they are protected from damage, and bare conductors are permitted where supported on insulators. For systems running at a higher voltage, open insulated conductors can be used where run on insulated supports and protected from damage. Bare conductors are permitted where run on insulated supports and as guarded against accidental contact up to the point of termination using methods in 110.27.
669.7. Warning Signs. Warning signs must be posted indicating the presence of bare conductors.
669.8. Disconnecting Means. A disconnecting means must be provided from each dc power supply, which can consist of removable links or conductors.
669.9Overcurrent Protection. DC conductors must have overcurrent protection in the form of fuses or circuit breakers, or a current sensing means that operates a disconnecting means, or some other approved means.
670.2. Definitions
Industrial machinery (machine) This definition is sufficiently broadly written that it should not be necessary to routinely revisit the definition as technology changes. The provisions do not apply to any machine or tool which is not normally used in a fixed location and can be carried from place to place by hand. The concept in this article is that the rules in this article and Art. 430 apply to the point of connection to the main terminals of the machine. On the load side of those terminals, including control circuit protocols and workspace rules, NFPA 79, Electrical Standard for Electrical Machinery takes over. This principle applies to machinery prepared as a unit by its manufacturer, but then disassembled for shipment and reassembled on-site. However, normal NEC provisions apply when various components are provided from different vendors on-site and then field assembled into a process. For example, motor controllers applied to an extensive conveyor system may or may not come directly under NEC provisions depending on whether they were furnished by the OEM as part of his engineered system.
670.3. Machine Nameplate Data. This information has been correlated with 409.110 with respect to the short-circuit current rating requirements. If over-current protection is provided as covered in 670.4(B), the machine must be marked accordingly.
670.4. Supply Conductors and Overcurrent Protection. Part (A) requires the supply wiring to have an ampacity not less than 125 percent of resistance loads, plus 125 percent of the largest motor plus 100 percent of all other motors and apparatus, based on their duty cycle and likelihood of simultaneous operation. Part (B) defines the machine as a single entity for the purposes of disconnecting rules, and therefore must be provided with a disconnecting means. Although not stated here, NFPA 79 is correlated with 404.8(A) and uses the same 2.0 m (6 ft 7 in.) height limit, measured in the same way (the center of the grip in its uppermost position), and arranged to be lockable in the open position. When so locked, the machine cannot be energized by any local or remote action. Part (C) reiterates the rules of 430.62 for motor feeder overcurrent protection limits, because that is, after all, the function included an overcurrent device on which one of these units performs.
675.1. Scope. This article covers electrically driven irrigation machines that are not portable by hand, and used primarily for agricultural purposes. It does not cover water pumps that bring water to the machines.
675.3. Irrigation Cable. This is the cable used to interconnect enclosures on an irrigation machine. It has a number of construction specifications predicated on its expected use in a wet location and where subject to flexing, and requiring metallic armoring in its inner construction. It must be supported at least every 1.2 m (4 ft). The cable is permitted to be a composite with control and grounding conductors included, and must be terminated in fittings designed for the cable.
675.7. Equivalent Current Ratings. For continuous duty the normal rules in Art. 430 apply. Where the duty cycle is inherently intermittent, Part (A) determines a continuous current rating by taking 125 percent of the largest motor (using the nameplate rating), adding the sum of all other motor nameplate ratings, and then multiplying the final summation by the maximum percent duty cycle for which they can operate. Part (B) determines an equivalent locked-rotor current by adding the locked-rotor currents of the two largest motors to the full-load current ratings of the remaining motors.
675.8. Disconnecting Means. Part (A) requires the main controller governing the complete machine to at least match the equivalent continuous current rating determined as above, or per 675.22(A) for center pivot machines, and a horse power rating taken out of the locked-rotor tables at the end of Art. 430 based on the equivalent locked rotor current determined above or by 675.22(B) for center pivot machines. If a listed molded case switch is used for the disconnecting means, it must have the ampere rating as described but it need not have a horse power rating.
Part (B) requires a main disconnect, which must include overcurrent protection, and can be at the main power connection to the machine or not over 15 m (50 ft) away, provided it is within view and readily accessible with a permanent lock-open feature. Its ratings must not be less than as required for the main controller, although it is recognized that circuit breakers do not carry horsepower ratings.
Part (C) requires individual disconnects for each motor and controller, as covered in Part IX of Art. 430. These disconnects need not be readily accessible.
675.9. Branch-Circuit Conductors. The branch-circuit conductor minimum ampacity follows the continuous current ratings determined by 675.7(A) or 675.22(A).
675.10. Several Motors on One Branch Circuit. This section varies the normal limits in 430.53, and allows several motors, each not over 2 hp and protected not over 30 A at not over 600 V, to operate off of taps made up with copper wire not smaller than 14 AWG and not longer than 7.5 m (25 ft). Each motor so connected must have normal running overload protection and its full-load current must be limited to 6 A. If these conditions apply, individual short-circuit and ground-fault protective devices need not be installed.
675.11. Collector Rings. Collector rings must carry their motor loads, as determined in one of three ways: 1) the result from 675.7(A); or 2) the result from 675.22(A); or 3) 125 percent of the largest controller (NEC says “device”; a motor is not a device but a controller is) plus the full load currents of all other “devices” served. The grounding ring must be fully sized to the same rating. Control and signaling rings must carry their load taken as 125 percent of the largest device plus 100 percent of all other devices. The rings must be in a suitable enclosure so they will withstand the environment they will operate within.
675.12. Grounding. All electrical equipment and enclosures on the machine must be grounded, although a machine that is electrically controlled by not electrically driven is exempt from grounding if the voltage is 30 V or less and the control circuits are power limited in accordance with Chap. 9, Tables 11(A) and 11(B). This last provision means that the control circuits operate within Class 2 or Class 3 parameters, but need not meet the source requirements for these circuits in 725.121(A) that would actually qualify them for this designation.
675.13. Methods of Grounding. This rule requires on-machine grounding conductors to be fully sized to the ungrounded conductors, and not as permitted in Table 250.122. However, feeder circuits are permitted to use the normal sizing rules.
675.15. Lightning Protection. Irrigation machines with a stationary point must be connected to a grounding electrode system as covered in Part III of Art. 250. Note that this connection alone will not meet the minimum requirements of NFPA 780, Standard for the Installation of Lightning Protection Systems, but it is the NEC minimum.
675.22. Equivalent Current Ratings. This section covers the calculations for center pivot machines that differ from other types of irrigation machinery. Part (A) determines a continuous current rating by taking 125 percent of the largest motor (using the nameplate rating) and adding 60 percent of the sum of all other motor nameplate ratings. Part (B) determines an equivalent locked-rotor current by adding double the locked-rotor currents of the largest motor to 80 percent of the full-load current ratings of the remaining motors.
680.1. Scope. Electrification of swimming, wading, therapeutic, and decorative pools, along with fountains, hot tubs, spas, and hydromassage bathtubs, has been the subject of extensive design and Code development over recent years. Details on circuit design and equipment layout are covered in NE Code Art. 680. Careful reference to this article should be made in connection with any design work on pools, fountains, and so forth.
Research work conducted by Underwriters Laboratories and others indicated that an electric shock could be received in two different ways. One of these involved the existence in the water of an electrical potential with respect to ground, and the other involved the existence of a potential gradient in the water itself.
A person standing in the pool and touching the energized enclosure of faulty equipment located at poolside would be subject to a severe electrical shock because of the good ground which his or her body would establish through the water and pool to earth. Accordingly, the provisions of this article specify construction and installation that can minimize hazards in and adjacent to pools and fountains.
The potential gradient in the water presents primarily a drowning hazard and not an electrocution hazard. It was determined by actual human volunteers immersed in the water, particularly with any water in their ears, that a gradient of as little as 4 V was enough to cause disorientation to the point that the individual may not be able to leave the water. This is why the bonding requirements for a hydromassage bathtub are less severe than for a spa or hot tub; the bathtub does not present the drowning hazard that a spa or hot tub, or larger pool presents.
Very important: Therapeutic pools in a health care facility are not exempt from this article. Therapeutic pools in hospitals are subject to all applicable rules in Art. 680, Part VI.
As noted at the end of the first sentence, the rules here also govern the installation of “metallic auxiliary equipment, such as pumps, filters, and similar equipment.” That wording has the effect of requiring that any such “metallic auxiliary equipment” satisfy the requirements of Art. 680. Where a particular installation detail addresses say, a circulating pump, the pump must be installed as described. This is very clearly stated by the first sentence.
One last note applies to the phrase body of water. Where a rule refers to a “body of water,” the rule applies to all types of pools, and the like, that are covered by Art. 680. Where individual types of pools, and the like, are identified, then the rule given there applies only to the specified types of pools, and the like.
680.2. Definitions. These definitions are important to correct, effective application of Code rules of Art. 680. Figure 680-1 shows a typical dry-niche swimming pool luminaire. Figure 680-2 shows a forming shell for a wet-niche luminaire.
The definition of cord-and-plug-connected lighting assembly covers a luminaire of all-plastic construction for use in the wall of a spa, hot tub, or storable pool. This type of luminaire operates from a cord-and-plug-connected transformer, and it does not require a metal niche around the luminaire.
Fig. 680-1. Dry-niche luminaire lights underwater area through glass “window.” (Sec 680.2).
Fig. 680-2. Forming shell is a support for the lamp assembly of a wet-niche luminaire. (Sec. 680.2.)
A hydromassage bathtub is a “whirlpool” or “Jacuzzi” bath for an individual bather, which is smaller than a hydromassage pool (spa or hot tub), but is covered in 680.70 through 680.74. It is designed to discharge its water after use.
The definition of maximum water level is, essentially, the deck level, because that is the level where the water can “spill out.” It is not, as generally previously supposed, the maximum fill level of the skimmer trough, because the water doesn’t actually spill out at that point. This definition has serious repercussions on the placement of swimming-pool junction boxes, as covered later.
The definition of no-niche luminaire covers a luminaire for installation above or below the water without any niche. This definition correlates to 680.23(D), which describes the installation of such luminaires.
Because there are differences in the requirements for “permanently installed” pools and “storable” pools, there has been some confusion in the past as to just what a “storable” pool is. A storable swimming or wading pool must not hold more than 1.07 m (3½) ft of water, or with plastic or inflatable walls of any dimension, and be “constructed on or above the ground.” The last part of this definition identifies specific types of pool construction that are considered to be storable regardless of water depth.
This definition, as well as two others (pool and permanently installed swimming, wading, immersion, and therapeutic pools) was changed to add the word “immersion” into the definition (pool), or into the title of the definition. The new title for this definition is “storable swimming, wading, immersion, and therapeutic pools.” The changes were made to clarify that a baptistery, whether permanently installed and maintained, or as part of a storable assembly, was covered by the applicable parts of Art. 680.
Figure 680-3 shows a “wet-niche luminaire.”
Fig. 680-3. Wet-niche luminaire consists of forming shell set in pool wall with cord-connected lamp-and-lens assembly that attaches to the forming shell, with cord coiled within the shell housing. (Sec. 680.2.)
680.5. Ground-Fault Circuit Interrupters. This section describes ground-fault circuit interrupters (GFCIs) that are required to be used by other rules of this article. Additional protection may be accomplished, even where not required, by the use of a GFCI. Since the GFCI operates on the principle of line-to-ground leaks or breakdowns, it senses, at low levels of magnitude and duration, any fault currents to ground caused by accidental contact with energized parts of electrical equipment. Because the ground-fault interrupter operates at a fraction of the current required to trip a 15-A CB, its presence is mandatory under certain Code rules and is generally very desirable. As indicated by the wording here, either receptacle, CB, or “other listed type” of GFCIs may be used to satisfy the rules of Art. 680 where a GFCI is required.
680.7. Cord-and-Plug-Connected Equipment. The 3-ft (900-mm) cord limitation mentioned in this rule would not apply to swimming pool filter pumps used with storable pools under part III of Art. 680, because these pumps are considered as portable instead of fixed or stationary. See the comments following 680.30. However, as covered by the UL White Book, pumps for permanently installed pools must be provided with a 3-ft (900-mm) cord and be factory-equipped with either a locking- or nonlocking-type attachment plug. Those equipped with the locking-type attachment plug are intended for use within the 5- to 10-ft (1.5- to 3.0-m) exclusion area, as described by 680.22(A)(1). Those equipped with nonlocking-type attachment plugs must be located at least 10 ft (3.0 m) from the inside walls of the pool, as called for by 680.22(A)(2). The equipment grounding conductor in the cord, which as a practical matter usually drives the overall cord size, must not be smaller than 12 AWG, or larger if required by 250.122.
680.8. Overhead Conductor Clearances. The general rule states that service drops and open overhead wiring must not be installed above a swimming pool or surrounding area extending 10 ft (3.0 m) horizontally from the pool edge, or diving structure, observation stands, towers, or platforms. But, the exceptions exempt only utility company lines from the rule, provided the designated clearances are satisfied.
Item C in Code Table 680.8 and the diagram clarify the horizontal dimensions around the pool to which the clearances of the table apply for utility lines over a pool area (Fig. 680-4, top). The dimension C, measured horizontally around a pool and its diving structure, establishes the area above which utility lines (and only utility lines) are permitted, provided the clearance dimensions of A or B in the table are observed. The dimensions A and B do not extend to the ground as radii, and the dimension C is the sole ruling factor on the “horizontal limit” of the area above which the clearances of A and B apply.
As the basic rule is worded—and the table and diagram specify—the clearances of A and B must be observed for utility lines above the water and above that area at least 10 ft (3.0 m) back from the edge of the pool, all around the pool. But the horizontal distance would have to be greater if any part of the diving structure extended back farther than 10 ft (3.0 m) from the pool’s edge. If, say, the diving structure extended back 14.5 ft (4.4 m) from the edge, then the overhead line clearances of A and B would be required above the area that extends 14.5 ft (4.4 m) back from the pool edge, not just 10 ft (3.0 m) back (Fig. 680-4, bottom).
Part C of the table says that the horizontal limit of the area over which the required vertical clearances apply extends to the “outer edge of the structures listed in (1) and (2).” That wording clearly excludes item 3 of the basic rule (observation stands) from the need to extend the horizontal limit over 10 ft (3.0 m), as shown at the bottom of Fig. 680-4.
The last paragraph in this section provides guidance on use of telephone company overhead lines and community antenna system cables above swimming pools. Although the first sentence of 680.8 generally prohibits “service-drop or other open overhead wiring” above pools, it was never the intent that the rules of this section apply to telephone lines. The general concept of this wording is to specifically permit such lines above pools provided that such conductors and their supporting messengers have a clearance of not less than 10 ft (3.0 m) above the pool and above diving structures and observation stands, towers, or platforms. However, network-powered broadband communications drops can operate at significantly higher voltages and must meet the normal clearances for the 0-750 V column in Table 680.8.
Fig. 680-4. Clearances from Table 680.8 apply as indicated in these diagrams. (Sec. 680.8.)
680.9. Electric Pool Water Heaters. A swimming pool heater requires branch-circuit conductor ampacity and rating of the CB or fuses at least equal to 125 percent of the nameplate load current. An electrically powered swimming pool heater is considered to be a continuous load and is therefore made subject to the same requirements given in 422.13 for hot water tanks. In addition, large units must meet the usual load subdivision rules, as for example in 424.22(B).
680.10. Underground Wiring Location. This section is aimed at eliminating the hazard that underground wiring can present under fault conditions that create high potential fields in the earth and in the deck adjacent to a pool. Aside from the electric circuits associated with pool equipment, underground wiring must not be run within the ground closer than 5 ft (1.5 m) from the inside of a pool. When inadequate space requires that extraneous underground circuits be run within the ground under the 5-ft (1.5-m) horizontal band around the pool, such wiring is permitted provided that (1) any such circuits are in “complete raceway systems” of rigid metal conduit, IMC, or rigid nonmetallic conduit, (2) the raceways are galvanized steel or otherwise provided with corrosion resistance, (3) the raceways are suitable for the location (by complying with underground application data from the UL’s Electrical Construction Materials Directory), and (4) the burial depths of the raceways conform to the table of burial depths, given in this section. The complete systems rule, new in the 2008 NEC, eliminates the possibility of simply sleeving a segment of a UF cable run in pipe, for example.
680.12. Maintenance Disconnecting Means. Here the Code mandates that a disconnect—one or more, as needed—must be provided to disconnect all ungrounded conductors for all pool-associated equipment other than lighting. This disconnecting means can be a unit switch where the switch disconnects all ungrounded conductors, or it can be a CB in a control panel, or other Code-recognized disconnecting means. If no such disconnect is available, then additional disconnecting means are required to be installed. The required maintenance disconnect must be “within sight” of the utilization equipment, which means not more than 50 ft (15 m) from the equipment, visible, and not guarded by a door that prevents direct access. It must be at least 1.5 m (5 ft) from the pool measured horizontally from the water’s edge, although it can be nearer if behind a permanent barrier that provides a reach path of at least that extent.
680.20. General. This rule mandates compliance with the provisions of parts I and II in this article where installing electrical and pool-associated equipment at a permanently installed swimming pool.
680.21. Wiring Methods. In part (A)(1), the rule specifically requires an equipment grounding conductor for “pool-associated motors.” The rule here requires that a circuit to a pool filter pump—or any other “pool-associated motor”—must be run in rigid metal conduit (steel or aluminum or red brass), in intermediate metal conduit (so-called IMC), in rigid nonmetallic conduit (such as Schedule 40 or Schedule 80 PVC conduit), or in Type MC cable that is “listed” for the application. And for all such circuits to pool motors, a separate equipment grounding conductor of the proper size must be run in the raceway or cable with the branch-circuit conductors. The equipment grounding conductor must be sized from NEC Table 250.122, based on the rating of the overcurrent protective device (the fuse or CB) protecting the branch-circuit wires. It must never be smaller than No. 12, insulated, and must always be copper. Note that clarifying language added to the parent section in the 2008 NEC clarifies that the rules in (1) apply to all motor installations, except as modified in particular circumstances in (2) through (5).
Note: This rule clearly eliminates past confusion and disputed Code practice. It requires one of the three raceways or Type MC cable to feed a filter pump and does not permit use of Type UF or type use cable for the pump circuit, as shown in Fig. 680-5.
Fig. 680-5. A very clear rule is now applied to wiring method required for pump. (Sec. 680.21.)
Part (A)(2) offers only limited use of EMT as part of the circuit to a pool motor. EMT may be used as part of the circuit where the raceway is within or on a building, but may not be used overhead or underground outdoors. Any part of the circuit outdoors (not on a building) must be one of the three rigid conduits described in the basic rule. Part (A)(3) recognizes the use of liquidtight flexible conduit as the supply raceway for a pool motor where flexibility is needed. Such liquidtight flex would have to be a metallic type, unless it is nonmetallic flex UL listed for outdoor use. But in all cases where EMT or liquidtight flex is used as permitted by the rules of parts (A)(2) and (A)(3), a separate equipment grounding conductor must be used, as previously described, within the raceway.
As permitted by part (A)(4), wiring to a pool-associated motor may be NM cable or any of the NEC wiring methods for that part of the circuit that is run in the interior of a one-family dwelling unit. But note that the last paragraph specifies that interior wiring of a one-family dwelling may be part of the circuit to a filter pump only if the interior circuit has at least a No. 12 insulated or covered ground wire. This would recognize Romex with a No. 12 ground wire, but BX (Type AC) with its No. 16 aluminum bonding strip would have to contain an additional insulated grounding conductor that is covered or insulated and at least a No. 12. This wording recognizes that wiring within a building is under better protection for the reliability of the equipment grounding conductor.
The last part, (A)(5), of this section permits use of flexible cord for cord-and-plug-connected pump motors, under the same conditions as described for general cord and plug connection in 680.7. The wording of the requirement here for sizing the equipment grounding conductor within the flexible cord would appear to permit the use of a No 14 copper, but that is not the case. The general wording in 680.21(A)(1) requires a 12 AWG minimum equipment ground, and the rules there apply to all motors unless modified in (2) through (5). Since this provision is not expressed as a modification, no such modification is allowed and the 12 AWG sizing minimum prevails.
The wording of 680.21(B) excludes cord-and-plug-connected double-insulated pumps from the need for bonding. Where such a pump is used, all other components of the pool and its associated equipment must be bonded together, but are required to connect to the pump’s equipment grounding conductor. Presumably this can be done at the receptacle outlet supplying the pump.
680.22. Area Lighting, Receptacles, and Equipment. The basic rule here prohibits receptacles within 10 ft (3.0 m) from the pool edge, and part (A)(4) calls for GFCIs to protect all 120-V receptacles located between 10 and 20 ft (3.0 and 6.0 m) of the inside walls of indoor and outdoor pools. But part (A)(1) permits the installation of a receptacle for a swimming pool or fountain recirculating pump, “or other loads directly related to circulation and sanitation,” less than 10 ft (3.0 m) but not closer than 6 ft (1.83 m) from the inside wall of the pool. Normally, receptacles are prohibited from installation anywhere within the 10-ft (3.0-m) boundary around the edge of the pool at dwelling units as covered in (A)(3). However, because swimming pool pump motors are commonly cord-connected to permit their removal during cold weather in areas where freezing may damage them, this rule applies to a receptacle for the pump motor. Such a receptacle must be a single receptacle of the locking and grounding type and must have GFCI protection.
Part (A)(2) allows other receptacles, of other configurations, voltages, etc., to be no closer than 1.83 m (6 ft) from the pool. This would include convenience receptacles at other than residential occupancies.
Part (A)(3) requires, for dwelling units, at least one 120-V receptacle to be “located” within the 10-to 20-ft (3.0- to 6.0-m) band around the pool for any permanent pool at a dwelling unit (e.g., a one-family house). The word located was put in to replace the word installed, because there is no need to install such a receptacle if there is already one located within that area around the pool. The rule requires a minimum of one 120-V receptacle at every pool installed at a dwelling unit. This rule ensures that a receptacle will be available at the pool location to provide for the use of cord-connected equipment. It was found that the absence of such a requirement resulted in excessive use of long extension cords to make power available for appliances and devices used at pool areas. This required receptacle may not be mounted more than 2.0 m (6½ ft) above the pool, and should be located so that it is visible after all the pool associated equipment is “hidden” by shrubs or other natural or man-made partitions.
Part (A)(4) requires that all receptacles within 20 ft (6.0 m) of the inside wall of the pool must be protected by a GFCI. Of course, if the pool is on the property of a private home, all outdoor receptacles must have GFCI protection—at any distance from the pool.
Part (A)(5) says measurement of the prescribed distances of a receptacle from a pool is made over an unobstructed route from the receptacle to the pool, with hinged or sliding doors, windows and walls, floors, and ceilings considered to be “effective permanent barriers.” If a receptacle is physically only, say, 3 ft (900 mm) from the edge of the pool but a hinged or sliding door is between the pool edge and the receptacle, then the distance from the receptacle to the pool is considered to be infinite, and the receptacle is thus more than 20 ft (6.0 m) from the inside wall of the pool and does not require GFCI protection (Fig. 680-6, bottom).
Part (B) requires that any pool circulation motor on 15 or 20 A circuits, single phase and on either 120 or 240 V circuits must incorporate GFCI protection for the motor outlet whether the motor is hard wired or plugged into a receptacle. This wording needs serious editorial assistance because it does not literally apply to a pool motor operating on two legs of a 208Y/120 V distribution. In addition, the entire rule is placed in the wrong section of the article. This rule needs to be relocated into Sec. 680.20 whose scope is the circulation motors addressed here.
The reference to “existing installations” in 680.22(C)(3) must be understood to refer to luminaires that are already in place on a building or structure or pole at the time construction of the pool begins. Where a pool is installed close to, say, a home or country club building, luminaires attached to the already existing structure may fall within the shaded area for a band of space 5 ft (1.5 m) wide, extending from 5 ft (1.5 m) above the water level to 12 ft (3.7 m) above water level all around the perimeter of the pool, as described in part (C)(1). The requirement for a lighting outlet so located to be provided with GFCI protection has been removed from this section on the basis that such protection is a negligible safety factor. However, new luminaires may not be installed in that space band around the pool.
Under the conditions given in part (C)(2), luminaires may be installed less than 12 ft (3.7 m) above the water of indoor pools. Luminaires that are totally enclosed and supplied by a circuit with GFCI protection may be installed where there is at least 7½ ft (2.3 m) of clearance between the maximum water level and the lowest part of the luminaire.
Fig. 680-6. Rules cover all receptacles within 20 ft (6.0 m) of the pool’s edge. (Sec. 680.22.)
Fig. 680-7. The luminaires over the pool and for 5 ft (1.5 m) back from the edge must be at least 12 ft (3.7 m) above the maximum water level if their supply circuit is without GFCI protection. If GFCI protection is provided by a GFCI-type circuit breaker in the supply circuit(s) to these totally enclosed luminaires, their mounting height may be reduced to a minimum of only 7½ ft (2.3 m) clearance above water level for an indoor (but not an outdoor) pool. The luminaires at right side do not require GFCI protection because they are over 5 ft (1.5 m) above the water level and rigidly attached to the structure. Refer to Fig. 680-8. [Sec. 680.22(C)(2) and (C)(4).]
Figure 680-7 applies the foregoing rules to lighting at an enclosed pool.
Part (D) requires that switching devices be at least 5 ft (1.5 m) from the pool’s edge or be guarded [Fig. 680-8(c)]. To eliminate possible shock hazard to persons in the water of a pool, all switching devices—toggle switches, CBs, safety switches, time switches, contactors, relays, and so on—must be at least 5 ft (1.5 m) back from the edge of the pool, or they must be behind a wall or barrier that will prevent a person in the pool from contacting them. However, where the switch is specifically listed for installation within the 5-ft (1.5-m) horizontal exclusion area, then the switch may be located closer than 5 ft (1.5 m), where listed for such installation, as indicated by the last sentence.
680.23. Underwater Luminaires. In part (A) of this section, the wording must be followed carefully to avoid confusion about the intent.
Part (1) starts by requiring that any underwater luminaire must be of such design as to ensure freedom from electric shock hazard when it is in use and must provide that protection without a GFCI. But in part (A)(3), a GFCI is required for all line-voltage luminaires (any operating over 15 V, such as a 120-V luminaire) to provide protection against shock hazard during relamping. A GFCI is not required for low-voltage swimming pool lights (12-V units).
The UL presents certain essential data on use of GFCI devices, which must be factored into application of such devices, as follows:
A GFCI is a device whose function is to interrupt the electric circuit to the load when a fault current to ground exceeds some predetermined value that is less than that required to operate the overcurrent protective device of the circuit.
Fig. 680-8. For luminaires and switching devices, installed locations are governed by space bands around pool perimeter. [Sec. 680.22(C).]
A GFCI is intended to be used only in circuits where one of the conductors is solidly grounded.
Class A GFCIs trip when the current to ground has a value in the range of 4 through 6 mA. Class A GFCIs are suitable for use in branch and feeder circuits, including swimming pool circuits. However, swimming pool circuits installed before local adoption of the 1965 NEC may include sufficient leakage current to cause a Class A GFCI to trip.
Class B GFCIs trip when the current to ground exceeds 20 mA. These devices are suitable for use with underwater swimming pool luminaires installed before the adoption of the 1965 NEC.
GFCIs of the enclosed type that have not been found suitable for use where they will be exposed to rain are so marked.
It should be noted that only a “listed” luminaire should be used—which means only a luminaire listed by UL or other test lab for use at a permanently installed pool, as called for by part (A)(8). In fact, the requirements given in parts (A)(5) through (A)(7) are very similar to UL data regarding the installation of listed underwater luminaires. UL data on listed luminaires must be carefully observed.
Part (4) sets 150 V as the maximum permitted for a pool luminaire, which means that the usual 120-V listed luminaires are acceptable. Note that the actual wording limits “supply circuits” which allows for a HID luminaire with a higher starting voltage, provided the line connections remained at 120 V.
Part (5) repeats the UL limitation on mounting distance of a luminaire below water level. When installed, the top edge of the luminaire must be at least 18 in. (450-mm) below the normal level of the pool water (Fig. 680-9). This 18-in. (450-mm) rule was adopted to keep the luminaire away from a person’s “chestarea,” because this is the vital area of the body concerning electric shocks in swimming pools. Keeping the top of the luminaire 18 in. (450 mm) below the normal water level avoids a swimmer’s chest area when he or she is hanging onto the edge of the pool while in the water. But, as the last sentence notes, an underwater luminaire may be used at less than 18 in. (450 mm) below the water surface if it is a unit that is identified for use at a depth of not less than 4 in (100 mm). Part (6) requires upward-facing luminaires to have their lens “adequately” protected, or be listed for use without a guard.
Fig. 680-9. Mounting of luminaire and circuit components must observe all Code rules and their specific dimensions. (Sec. 680.23.) Note that with the definition of maximum water level now at the point where the water spills out, the water height and the deck height are one and the same in most cases, and therefore the swimming pool junction box will be 200 mm (8 in.) above the deck. Also note that only rigid (or intermediate) metal conduit can support a box directly, without additional support direct to structure, per 314.23(E). Therefore, unless both conduits are metal [two is the minimum for box support in 314.23(E)] additional direct support is required for this swimming pool junction box.
Part (7) presents an interesting requirement on the use of wet-niche luminaires. The rule here requires that some type of cutoff or other inherent means be provided to protect against overheating of wet-niche luminaires that are not submerged but are types that depend on submersion in water for their safe operation. Note that the UL rules quoted here require some luminaires to be marked “Submerse Before Lighting.” Manufacturers of such luminaires should incorporate this protection—such as in the form of a bimetal switch similar to those used in motor end-bells for motor overload protection.
Part (B) details the use of wet-niche luminaires. A wet-niche underwater lighting assembly consists of two parts: a forming shell, which is a metal structure designed to support a wet-niche luminaire in the pool wall, and a luminaire, which usually consists of a lamp within a housing furnished with a waterproof flexible cord and a sealed lens that is removable for relamping.
Part (B)(2) requires that the conduit between the forming shell and the junction box or transformer enclosure must be approved (1) rigid metal conduit or IMC and made of brass or other approved corrosion-resistant metal, or (2) liquidtight flexible nonmetallic or rigid nonmetallic conduit with a No. 8 insulated copper conductor installed in the conduit and connected to the junction box or transformer enclosure and to the forming shell enclosures. The No. 8 insulated copper wire may be stranded or solid. Each enclosure—the forming shell as well as the box—must contain approved grounding terminals.
Note the term “approved” with respect to metal conduit. This is a deliberate Chap. 6 amendment of a Chap. 3 rule, in this case 344.6, that requires all rigid metal conduit to be listed. Listed red brass conduit has not been generally available for decades. However, as heavy wall brass water pipe, it is available from some plumbing supply houses. It threads very well with the usual NPT threading dies, and is an extremely robust product. Without very heavy foot pressure on the bender shoe, an attempt to bend this product by pulling on the bender handle will bend the steel handle and not the pipe. The inside of this pipe is smoother than listed steel heavy wall conduits. The wording in this section at least makes this an option, subject to inspectional approval.
Figure 680-10 shows a typical connection from a forming shell to a transformer enclosure supplying the 12-V lamp in the luminaire. If a 120-V luminaire is used, the conduit from the shell terminates in a junction box, as shown in Fig. 680-9. In the drawing of Fig. 680-10, from the forming shell, a length of metric designator 27 (trade size 1) PVC conduit extends directly to a 120/12-V transformer mounted on the back wall of a planter adjoining the pool [observing the 4-ft (1.2-m) back and 8-in.-high (200-mm-high) provisions of 680.24(B)(2)]. Where the nonmetallic conduit stubs up out of the planter soil, an LB (could be nonmetallic) connects the conduit to the transformer. The required 8 AWG conductor in the PVC conduit is terminated at the grounding bar in the transformer enclosure and on the inside terminal of an inside/outside grounding/bonding terminal on the forming shell. The external bonding lug provides for connecting the forming shell to the common bonding grid, as required by 680.26(A) and (B). The 8 AWG in the PVC conduit bonds the forming shell up to the transformer enclosure. Note that this 8 AWG conductor is not needed if metal conduit connects the shell to the transformer enclosure. One of the 12 AWG conductors in the supply circuit is an equipment grounding conductor that runs back to the panelboard grounding block and thereby grounds the 8 AWG and the metal fittings and transformer enclosure.
Fig. 680-10. Grounding and bonding is required in a typical hookup of low-voltage wet-niche luminaire with PVC conduit. (Sec. 680.23.) Note that in order to make the direct connection between the transformer and the luminaire as shown here, the transformer must be listed for such a direct connection. Further, it is to be presumed that such enclosures are not suitable for this task, unless they are specifically marked otherwise.
Part (B)(2)(b) requires that the inside forming shell termination of the No. 8 be covered with, or encapsulated in, a UL-listed potting compound. Experience has shown that corrosion occurs when connections are exposed to pool water. Listed epoxies are available to achieve this protection; however, some inspection agencies do accept a waterproof, permanently pliable silicone caulk compound.
Note that the illustrated assembly includes three non-current-carrying conductors: (1) an 8 AWG bonding conductor connecting the forming shell to the bonding grid; (2) an 8 AWG insulated conductor in PVC conduit between the forming shell and the transformer enclosure; and (3) a grounding conductor in the luminaire flexible supply cord. This is why 680.24(D) requires these enclosures to have at least one more grounding terminal than the number of conduit entries; with nonmetallic wiring, the run to the luminaire always has two grounding conductors.
Part (B)(4) requires sealing of the luminaire cord end and terminals within the wet-niche to prevent water from entering the luminaire. And grounding terminations must also be protected by potting compounds.
Part (B)(5) states that an underwater luminaire must be secured and grounded to the forming shell by a positive locking device which will ensure a low-resistance contact and require a tool to remove the luminaire from the forming shell. This provides added assurance that luminaires will remain grounded because, in the case of wet-niche luminaires, the metal forming shell provides a bond between the raceway (or No. 8 conductor in PVC) connected to the forming shell and the non-current-carrying metal parts of the luminaire.
Part (B)(6) calls for the luminaires to be so located that they can be maintained or re-lamped without the need to get wet. That is, consideration must be given to ensure that such “servicing” activities can be performed from a “dry location.” Further, the relamping location must be accessible without going into the water. In some instances, including luminaires in the bottom of a pool, these rules may require extensive lengths of cord, perhaps longer than what could be accommodated in the wet niche, and this will require careful design planning and consultation with the owners.
Part (C) permits use of an approved dry-niche luminaire that may be installed outside the walls of the pool in closed recesses which are adequately drained and accessible for maintenance. For a dry-niche luminaire, a “deck box,” set in the concrete deck around the pool, may be used and fed by metal (rigid or IMC) or nonmetallic conduit from the service equipment or from a panelboard. Where the circuit conductors to the luminaire are run on or within a building, the exception to the rule permits the conductors to be enclosed in EMT—but rigid metal or IMC or rigid nonmetallic conduit must be used outdoors when not on a building. And such a deck box does not have to be 4 in. (100 mm) up and 4 ft (1.2 m) back from the pool edge, as required for a junction box for a wet-niche luminaire. (See Fig. 680-1.)
Some approved dry-niche luminaires are provided with an integral flush deck box used to change lamps. Such luminaires have a drain connection at the bottom of the luminaire to prevent accumulation of water or moisture.
Part (D) covers no-niche luminaires. They follow the same rules as wet-niche luminaires, except where a niche is mentioned, the no-niche bracket replaces it.
Part (E) covers a through-wall lighting assembly which is designed for above-grade use. Here again the rules in 680.23 apply, except in this case where the connecting point to the wiring system is specified, in this case the conduit hub on the luminaire replaces the niche or the bracket. Review the UL data on this luminaire, quoted in this book at 680.33 below. These luminaires are used for both storable and permanent pools, and frequently wired improperly.
Part (F) covers branch circuit wiring. that runs on the supply side of transformer enclosures and junction boxes as shown in Figs. 680-9 and 680-10, and on the supply side of dry-niche wiring compartments. The first paragraph covers permitted wiring methods, which include both metallic and nonmetallic heavy wall conduits (RMC, IMC, PVC, RTRC) along with liquidtight flexible nonmetallic conduit, which in the earth must be listed for burial as discussed in Art. 356. An exception follows that permits liquidtight flexible metal conduit to make connections to pool-lighting transformers in lengths up to 1.8 m (6 ft) for any single use and up to 3 m (10 ft) for the total run. The exception also covers liquidtight flexible nonmetallic conduit, but that is an error because the exception is merely permissive and this wiring method is already permitted without limitations in the main rule. In this regard the exception reiterates the provision already covered in 356.10(5) relative to the type of LFNC allowed in lengths over 1.8 m (6 ft). For wiring on buildings, EMT is added to the list, and for wiring in buildings the list grows again by adding ENT and Types MC and AC cables.
In all instances including the cabled methods an insulated 12 AWG (or larger if so required by 250.122) equipment grounding conductor must be installed. The exception that follows permits an equipment grounding conductor to remain sized in accordance with the branch-circuit protection where run between a transformer enclosure and a swimming pool junction box. Often the wire sizes are increased on transformer secondaries because of the higher currents that follow from the transformation down to 12 V.
The second paragraph adds joint and termination rules for grounding conductors. The basic rule requires the grounding conductor to be installed unbroken at any point. There are two locations, and only two, where this is not so. The first is where the same branch circuit daisy chains through swimming pool junction boxes, transformer enclosures, or dry-niche wiring compartments. In every instance, the connections must be on “grounding terminals.” Do not use twist-on wire connectors to make these joints. The second location for joints in the grounding conductor is at the one or more points where the luminaire(s) are controlled by various means including a simple snap switch. Here again, the connections must be on “grounding terminals.” This applies even in the case of snap switches, which means field installing a short segment of equipment grounding bus in the device box or some comparable method.
680.24. Junction Boxes and Enclosures for Transformers or Ground-Fault Circuit Interrupters. Part (A) covers junction boxes that connect to a conduit that extends directly to a pool-lighting forming shell or no-niche luminaire mounting bracket, such as shown in Fig. 680-9. The junction box must be of corrosion-resistant material provided with threaded hubs for the connections of conduit, or nonmetallic hubs for nonmetallic wiring methods. Part (A)(3) requires these boxes to ensure continuity between metal conduit entries, even if the box is plastic, through the use of an “integral” means. This would mean a box constructed to meet 314.3 Exception No. 2, (refer to the discussion at that point) and it is unlikely such a box will be encountered. Plan on using metal-based swimming pool junction boxes if you are using metal conduits for wiring.
For line-voltage (120-V) pool luminaires, the so-called deck box (set in the concrete deck around the pool) is no longer permissible (except where approved dry-niche luminaires include flush boxes as part of an approved assembly), because the deck box, which was installed flush in the concrete adjacent to the pool, was the major source of failure of branch-circuit, grounding, and luminaire conductors due to water accumulation within them. The rule of part (A)(2), covering low-voltage and line-voltage lighting, states that these junction boxes must be located not less than 4 in. (100 mm) above the ground level or above the pool deck, and not less than 8 in. (200 mm) above the maximum pool water level (whichever provides the greatest elevation), and not less than 4 ft (1.2 m) back from the pool perimeter.
Watch out for this placement issue on junction boxes. Now that the “maximum water level” has been defined as where the water can spill out, these two dimensions will use the same starting point (the deck) in most cases, thereby adding 100 mm (4 in.) to the elevation of most deck boxes. Swimming pool junction boxes require a domed cover to lift up and off the enclosed wiring, so a 200 mm (8 in.) spacing to the bottom translates to about 350 mm (14 in.) of minimum clearance from the deck to the underside of a diving board or other structure the box may be placed under. This will often complicate the location of these boxes, which are a tripping hazard if left in the open. This was the reason for the 100 mm (4-in.) height reduction when it first went into the NEC; with the new definition, it no longer has much practical effect.
The wording of part (A)(2) does make clear that the elevated junction box could be less than 4 ft (1.2 m) from the pool’s edge if a fence or wall were constructed around the pool, with the box on the side of the wall away from the pool, isolating the box from contact by a person in the pool. Or the box could be within 4 ft (1.2 m) of the edge if the box were on the other side of a permanent nonconductive barrier.
Important: The last part of the rule in 680.24(A)(2)(c) still permits flush deck boxes where underwater lighting systems are 15 V or less if approved potting compound is used in the deck boxes and the deck boxes are located 4 ft (1.2 m) from the edge of the pool. In Fig. 680-11, a deck box for a 12-V luminaire could be used in the deck but the use of the box less than 4 ft (1.2 m) from the pool’s edge might be considered a violation, which does not recognize the fence along the pool in the same way as the first part in part (A)(2)(b). That is, the fence is not mentioned in the first part as sufficient isolation of the box from the pool—although the installation certainly does comply with the basic concept in part (A)(2)(b).
Fig. 680-11. The fence here permits the box to be closer than 4 ft (1.2 m) from pool’s edge. (Sec. 680.24.)
Part (B) covers installation of enclosures for 12-V lighting transformers and for GFCIs that are required for line-voltage luminaires. Such enclosures may be installed indoors or at the pool location. If a ground-fault interrupter is utilized at a pool, its enclosure must be located not less than 4 ft (1.2 m) from the perimeter of the pool, unless separated by a permanent means, and must be elevated not less than 8 in. (200 mm), measured from the inside bottom of the box down to the pool deck or maximum water level, whichever provides higher mounting. These rules cover installation of transformer or GFCI enclosures that connect to a conduit that “extends directly” to a forming shell.
Part (B)(1) specifically and clearly mandates that “other enclosures” be listed and labeled for the purpose. Use of any enclosures for GFCIs or transformers permanently installed in swimming pools that are not listed specifically for use at swimming pools is a violation. Enclosures that are so listed will be engineered to meet the following requirements.
Part (B)(1)(1) requires any such enclosure connected to a conduit that extends directly to an underwater pool-light forming shell to have threaded hubs or bosses or a nonmetallic hub. An enclosure of cast construction with raised, threaded hubs or with threaded openings in the enclosure wall would satisfy that rule. But because approved swimming pool transformers are usually available only in sheet metal enclosures with knockouts, this is usually not practical. Such a transformer connection will not meet other requirements in this section and violate the UL rule that “unless marked otherwise, these transformers are not suitable for connection to a conduit that extends directly to a wet-niche or no-niche luminaire.” In general, plan on running a raceway between the transformer enclosure and a listed swimming pool junction box that will receive the conduit running to the luminaire.
Part (B)(1)(2) requires corrosion resistant construction, using brass, copper, stainless steel, plastic, comparable material. Often a combination of such materials are used.
Part (B)(1)(3) also requires that transformer or GFCI enclosures be provided with an approval seal (such as duct seal) at conduit connections to prevent circulation of air between the conduit and the enclosure; that they must have electrical continuity between every connected metal conduit and the grounding terminals by means of copper, brass, or other approved corrosion-resistant metal that is integral with the enclosures; that they must be located not less than 4 ft (1.2 m) from the inside walls of the pool (unless separated by a solid fence, wall, or other permanent barrier); and that they must be located not less than 8 in. (200 mm) from the ground level, pool deck, or maximum pool water level, whichever provides the greatest elevation. This distance is measured from the inside bottom of the enclosure. (See Fig. 680-10.)
Note that part (B)(1)(4) intends to ensure a grounding path from the enclosure and its grounding terminals to any metal conduit. The section specifically states “metal conduit.” Where PVC conduit is used, the provision is not applicable, and the No. 8 ground wire in the PVC bonds to the forming shell. However, the section requires electrical continuity between an enclosure and “every connected metal conduit.” The conduit feeding the transformer primary does not seem to be involved with that rule because the concern is with the grounding path between the transformer or GFCI enclosure and the forming shell and because the No. 12 equipment grounding conductor in the primary supply will carry any current from a fault originating within the transformer enclosure. Local Code authorities should be consulted on the point.
The phrase “integral with the enclosures” is meant to cover a situation where the enclosure is nonmetallic. In this case, electrical continuity between the metal conduits and the grounding terminals must be provided by one of the metals specified, and this “jumper” must be permanently attached to the non-metallic box so that it is “integral.” Refer to the discussion on the same topic at (A)(3) for swimming pool junction boxes; this is not a practical alternative.
In Fig. 680-10, the transformer enclosure is being used as a junction box to an underwater light, with the equipment grounding conductors terminated at the grounding bar and carried through. The figure is predicated on the unlikely (but conceivable) assumption that the transformer is listed for a direct connection to a forming shell, and therefore meets all the requirements for such enclosures. However, the primary purpose of this enclosure is to house the transformer. Parts (B), (C), (D), (E), and (F) of 680.24 still apply. 680.24(A) would apply to boxes connected directly to underwater lights and is intended to cover situations where splices, terminations, or pulling of conductors might be required. Again, plan on providing a listed swimming pool junction box, which will be designed to meet all these rules, at every conduit termination that runs to a forming shell.
Part (C) of this section warns against creating a tripping hazard or exposing enclosures to damage where they are elevated as required. It is also important to remember that these junction boxes must be afforded additional protection against damage if located on the walkway around the pool. For protection against impact, they may be installed under a diving board or adjacent to a permanent structure such as a lamppost or service pole. As noted in the discussion at (A)(2), panel action on the maximum water level definition has greatly complicated the placement of these boxes under many diving boards.
Part (D) can be satisfied simply by using listed equipment; this is the rule requiring one more terminal than conduit entries, as discussed at 680.23(B)(2)(b). Part (E) calls for strain relief to be added to the flexible cord of a wet-niche lighting luminaire at the termination of the cord within a junction box, a transformer enclosure, or a GFCI. This mechanism will be furnished with or as an integral part of any listed swimming pool junction box.
680.25. Feeders. The permitted wiring methods permitted for feeders to panels in cabanas or other locations that supply branch circuits covered by Art. 680 consist of heavy-wall metal and nonmetallic conduits (RMC except aluminum in the pool area, IMC, PVC, RTRC) and liquidtight flexible nonmetallic conduit. EMT is permitted on or within buildings. Existing feeders in cabled wiring are grandfathered with a crucial limitation: the cable assembly must include an equipment grounding conductor, and that conductor must meet 250.24(A)(5).
This means that if an existing subpanel has a regrounded neutral, that subpanel is off limits for swimming pool circuits. Regrounded neutrals elevate the potential to ground of all equipment grounding conductors connected to them with the amount depending on the impedance of the neutral and the amount of load. That voltage elevation is not tolerable in a swimming pool environment.
Part (B) requires an insulated equipment grounding conductor to run with the feeder conductors, sized according to the usual rules. An uninsulated equipment grounding conductor in a grandfathered cable assembly is still OK. For a separately derived system, it must be no smaller than 8 AWG or as given in Table 250.66 for large sources; this provision essentially agrees with 250.30(A)(2) although it lacks the upper sizing provisions of 250.102(C) and should probably point there or be removed. Note also that an existing feeder supplying existing premises wiring to an outbuilding with a regrounded neutral and local grounding electrode in accordance with 250.32(B) Exception is permitted to supply a swimming pool panel.
680.26. Equipotential Bonding. The 8 AWG conductor and bonding system required in this section are not part of the effective fault current path that the equipment grounding system establishes, although they are ultimately connected to it. When all the required bonding connections are made at a pool, the entire interconnected hookup will be grounded by the “equipment grounding conductor” that is required to be run to the filter pump and to the lighting junction boxes and is connected to the 8 AWG bonding conductor at the pump and in the boxes. In a pool without underwater lighting, the equipment grounding conductor run with a pump-motor circuit will be the sole grounding connection for the bonded parts—and that is all that is required [see, however, 680.26(B)(6)]. The 8 AWG bonding conductor does not have to be run to a panelboard, service equipment ground block, or grounding electrode.
Its sole function is to reduce voltage gradients in the pool vicinity. A pool that is 750 V to ground because of a voltage transient is only a danger to a swimmer if one side is 752 V and the other side is 748 V, resulting in a 4 V gradient that could result in drowning, as noted in the discussion at the beginning of Art. 680. A bonding conductor extended to a remote panelboard in a basement could definitely save the life of an alien being with indefinitely long and stretchable arms who decided to reach through a cellar bulkhead and touch the panel while swimming in the pool. For the rest of us such a connection is useless. Likewise, the bonding connections are irrelevant to whether the pool is connected to one or more grounding electrodes. The frequently imposed, and bogus, requirement to drive ground rods around a pool perimeter may assist users in avoiding shocks from the adjacent azalea bushes, but nothing else.
The 8 AWG conductor mandated in this section is a bonding conductor, not an equipment grounding conductor. It must be solid in order to better withstand chemical attack from pool chemicals. Certain controversies regarding this wire in terms of solid versus stranded, insulated or not, color coded or not are resolvable if its status as a bonding conductor is kept in mind.
For example, no rule requires it to be insulated. If it is insulated, inspectors might require green color coding at any permanently exposed termination, if a rigid interpretation is put on 250.119 and if the “bonding” conductor is considered to be an “equipment grounding” conductor. That view is incorrect because 250.102 on “bonding jumpers” does not specify color of insulation or covering any more than there is a color code for a grounding electrode conductor.
Conflict has also arisen in the past over use of this 8 AWG bonding conductor because 310.3 requires 8 AWG and larger conductors to be stranded where installed in raceways. That rule had the effect of limiting use of solid No. 8 conductors, and their manufacture seemed to be curtailed. 310.3 has an exception which exempts conductors from being stranded in raceways where other Code rules permit such application. However, while the “8 AWG equipment grounding conductor” covered by 680.23(B)(2)(b) is permitted to be either stranded or solid, the 8 AWG that serves as the common bonding grid is not so exempt and must always be solid.
Part (A) gives a general statement aimed at defining the nature and performance of the bonding connection. This connection is not intended to act as an equipment grounding conductor, which, of course, carries fault current to ensure operation of the circuit OC device in the case of a faulted conductor. The bonding connection required here is only intended to ensure that all non-current-carrying parts of the pool and its electrical system are at the same potential with respect to ground at the pool location, thereby reducing voltage gradients in and around the pool (Fig. 680-12). Note that the bonding requirements for components required as part of the bonding grid are not necessarily dependent on any particular proximity to the pool, such as 1.5 m (5 ft). For example, a recirculating pump motor for a pool and the associated piping functionally extends the pool water zone, and requires bonding however far it is from the pool.
Part (B) reiterates the concept that bonding conductors need not run to remote panels because that would be irrelevant to their intended function. It then spells out in detail the pool components that must be bonded together, using 8 AWG solid copper wire. It also mentions brass conduit as a bonding conductor, which is incomprehensible in terms of functioning as a wire, but apparently intended to address a brass conduit as a collection point for wired bonding connections. Figure 680-13 shows what could be an example. This would be a reasonable application of the wording. There are seven categories of components that must be included in the bonding connections (Fig. 680-14).
Fig. 680-12. All of these metallic, non-current-carrying parts of a pool installation must be “bonded together” through the “equipotential bonding grid.” (Sec. 680.26.)
Paragraph (1) covers conductive pool shells. These include pools made of concrete, including concrete block, whether or not it is supported by steel reinforcing. Bonding must be accomplished to the reinforcing steel using bonding connections that are listed for direct burial and for rebar connections; refer to the discussion in this book at 250.52(A)(3) (concrete-encased electrodes) for more information on this point. The reinforcing steel used for this purpose must be the usual, unencapsulated product with crossing points tied by steel tie wires or equivalent methods. If the reinforcing steel is the epoxy coated (typically green) product commonly used for highway bridge construction, then it is ignored and an alternative method is substituted.
Fig. 680-13. 8 AWG insulated bonding conductors from several bonded items connect to the brass conduit, thereby using it as a bonding conductor. 680.26(B)(7) requires bonding of all “metal conduit” within 5 ft (1.5 m) of pool edge. (Sec. 680.26.)
Fig. 680-14. All designated parts must be connected to an equipotential bonding grid. (Sec. 680.26.)
The Code does not require each individual reinforcing bar to be bonded. It recognizes that the steel tie wires used to secure the rebars together where they cross each other provide the required bonding of the individual rods. Tests conducted over a period of several years by the NE Code Technical Subcommittee on Swimming Pools have shown the resistance of the path from one end to the other through the structural steel to remain at less than 0.001 
If conventional steel reinforcing is not available on a pool defined as having a “conductive pool shell,” then a mandatory alternative method applies. Warning: This method is so expensive that the cost of the required bonding grin alone is likely to extremely costly, especially after considering required changes to normal trade scheduling in order to accommodate the additional work. Try to make customers aware, as early in the design process as possible, that the lack of conventional reinforcing steel may have this effect. The alternative requires a copper bonding grid to cocoon the pool structure, below as well as on the sides, and arranged to approximate the coverage of normal steel reinforcing. Specifically, 8 AWG bare sold wires must envelop the pool at a distance of no more than 150 mm (6 in.) using a 300 mm by 300 mm (1 ft by 1 ft) pattern (with a tolerance of 100 mm or 4 in.), and with all points of crossing connected in a manner suitable for direct burial. The only practical way to cope with this is through the use of substation grounding mats. These are available in 6-ft wide rolls with a range of bare copper wires and welded grid patterns to choose from, including the 8 AWG 300 mm (1 ft) pattern mandated here.
Paragraph (2) covers the pool deck and any other areas that immediately surround the pool perimeter. If there is steel reinforcing in the deck, then it gets the same bonding treatment as for the walls of a pool that use steel reinforcing. The deck reinforcing must be bonded to the method used for the pool walls at no fewer than four points evenly distributed around the edge of the pool. If there is no steel reinforcing in the deck, or if it is epoxy coated, then an alternate method is provided. Unlike the alternate method for the pool walls, the alternate for the deck reinforcing is simple, inexpensive, practical, and well-rooted in historical pool bonding practice. A 8 AWG bare solid copper wire is run around the pool at a distance of 450 to 600 mm (18 in. to 24 in.) from the inside of the pool wall. It must be buried 100 mm to 150 mm (4 in. to 6 in.) “below the subgrade” which presumably means the distance below the final surface treatment such as coping stones. Any spliced must be listed for the application, which means listed and marked “DB” or “direct burial” and comprised of all copper, brass, or bronze components, or of stainless steel.
Part (3) covers metallic components of the pool structure not already covered. This would include the walls of metal-walled pools (Fig. 680-15). As before, epoxy-coated reinforcing steel is ignored; thereby avoiding a conflict with 680.26(B)(1)(a). The rule does not require individual sections of such pools to be bonded. However, the overlapping ends of each section to be bolted must not be painted. If they are, the paint must be removed completely to restore conductivity. In addition, resistance tests should be made across each bolted section after assembly to ensure low resistance. These sections normally are fastened together by corrosion-resistant bolts at least 
in. (9.5 mm) in diameter, and such an installation satisfies the bonding objectives. But electrical parts of such a pool must be tied into that common bonding grid by 8 AWG bonding conductors.
Fig. 680-15. Walls of bolted or welded metal pool are metallic parts of the pool structure and require bonding. (Sec. 680.26.)
Part (4) covers the back side of metal forming shells and no-niche luminaire brackets. An exception exempts listed luminaire forming shells that are part of a low-voltage lighting system that does not require bonding.
Part (5) covers all metal fittings in or attached to the pool structure (Fig. 680-16), with the usual exemption for small hand grips, etc., that are not over 100 mm (4 in.) in any dimension and that do not extend over 25 mm (1 in.) into the pool wall.
Fig. 680-16. Fittings for pool ladders have bonding strips attached to them for connection of the No. 8 bonding conductor. When the supports are set in the concrete deck, the bonding connections tie them all together. The 8 AWG bonding connections are shown (arrow) and a protective coating is painted on the connectors to protect against corrosion. (Sec. 680.26.)
Part (6) covers electrical equipment associated with the pool, including pump motors and the associated circulation system (Fig. 680-17) and conductive surfaces that are part of pool covers and their motors. Double insulated equipment does not, however, require incorporation into the bonding grid. This includes double-insulated pump motors, however, for those applications a run of 8 AWG bare copper must be extended from the bonding grid to the pump location, and left coiled there long enough so that if the double-insulated motor is replaced with a conventional motor, a bonding connection will be readily accessible.
Fig. 680-17. Water-fill pipe and metal housings associated with the water-circulation system are “bonded” with an insulated 8 AWG solid copper conductor, as required. (Sec. 680.26.)
In addition, in cases where no component in the bonding grid has a connection to the equipment grounding system for the premises, such a connection will be made at this point using this bonding conductor. This could be achieved by connecting it at the outlet passing through a wet-location extension box, or for hard-wired applications entering the raceway and extending to an enclosure through a tee conduit body using a suitable watertight gland connector, or any other workmanlike solution appropriate for the location.
This part also covers pool water heaters that generally require bonding. However, for heaters rated above 50 A and with specialized instructions, only the specifically designated parts are to be bonded, or grounded, or both.
Part (7) includes metal raceways, metal-clad cable assemblies, metal plumbing, and all other “fixed metal parts.” This also includes swimming pool junction boxes (Fig. 680-18) and swimming pool transformer enclosures, as well as metal fences, etc., if not separated by a barrier or more than 1.5 m (5 ft) away, measured horizontally from the pool walls. In addition, parts more than 3.7 m (12 ft) above the pool or observation stands, diving platforms, etc., are not required to be bonded. Figure 680-19 shows an example of the entire system put together.
In that drawing, the steel reinforcing rods, tied together with steel tie wires at intersections, are used as a common equipotential bonding grid to bond together pool equipment. Equipment shown here is required to be bonded. In addition, any metal parts (lighting standards, pipes, etc.) within 5 ft (1.5 m) of the inside walls of the pool and not separated from the pool by a permanent barrier must be bonded. All connections made must be completed with proper connectors, lugs, and so on, as required in 250.8 and as called for by the parent language in Part (B) of this section.
Fig. 680-18. Elevated metal junction box within 5 ft (1.52 m) of pool edge must be bonded. (Sec. 680.26.) Note that this is not a swimming pool junction box and therefore violates numerous provisions of 680.24(A)(1). However, the bonding is correct.
Part (C) of this section is unfortunately worded, because the terminology “bonded water” is an oxymoron, like “married bachelor.” Nevertheless reasonable substantiation was presented that there was a potential problem if there were no electrical connection whatsoever between the conductive pool water and the surrounding bonding grid, because there are significant issues with potential water-to-deck voltages. In past years, this would have been unheard of because something in the pool structure was always tied to the surrounding bonding grid. Today this is no longer an academic concern. Many pools are now being constructed with all nonmetallic framing, nonmetallic liners, double-insulated pump motors, nonmetallic ladders, no diving board and no in-pool lighting. This rule ensures that in such cases, there will be a functional contact between the water and the surrounding grid.
Nevertheless, there are issues with the requirement, even beyond the terminology about bonding to that which cannot be bonded. The rule is missing any provisions about corrosion resistance of the connecting medium, whether it will likely be disturbed during swimming, etc. It is probably the most idiotic application of a soft metric conversion in the entire NEC (5806 mm2 as if those 6 mm2 are going to make a difference). All of this will need to be sorted out with the inspector until the next code cycle. There are also practical issues of not wanting an 8 AWG conductor draped over the corner of a pool.
One way to address this in a very simple, workmanlike way is to put a current collector in the drain line to the pool, as close to the pool drain as possible. If the drain is a metric designator 53 (trade size 2) pipe, a brass or stainless steel nipple (check with the pool contractor for an assessment/preference as to reactivity with his chemicals) in the drain line. The area exposed on the inside of this nipple is the circumference multiplied by the length. A nipple just 50 mm (2 in.) long will expose more than enough area to the water.
Fig. 680-19. An 8 AWG bonding jumper ties each of the indicated parts to the rebar grid, completing the bonding. (Sec. 680.26.)
A = 
Dh = 3.14 × 2 in. × 2 in. = 12.5 in.2
Attach a 8 AWG solid bonding wire to this current collector using a pipe clamp listed for this pipe size, and further manufactured using all copper, brass, or bronze, or stainless steel components, enabling a further listing for direct burial and with the appropriate markings. This is critical because replacing this item would involve extensive excavation. Connect the other end of the bonding conductor to the bonding grid, using a direct burial rated splicing device at the other end. Note that this will not only connect the water to the bonding grid, but to the equipment grounding system for the premises as well. This is because, if the installation is fully compliant, the connection mandated in 680.26(B)(6)(a) will have been made as well.
680.27. Specialized Pool Equipment. Part (A) of this section treats connection of loudspeakers for underwater audio output in the same way as a wet-niche pool luminaire. Wording and rules are almost identical to those in 680.23(B)(5). Connection from the speaker forming shell is made to a junction box installed as set forth in 680.21(A) for a luminaire. The wiring methods correlate with 680.23(B)(2).
Part (B) covers pool covers, and the control location must be such that the operator can see the entire pool. The motor(s) must be located to provide at least 1.5 m (5 ft) of space between them and the pool wall, and they must have GFCI protection.
Part (C) covers deck area heating. These rules cover safe application of unit and radiant heaters. Such units must be securely installed, must be kept at least 5 ft (1.5 m) back from the edge of the pool, and must not be installed over a pool. Permanently wired radiant heaters have the same horizontal clearance, but add a 3.7 m (12-ft) vertical clearance as well. Radiant heating cables are prohibited from use embedded in the concrete deck.
680.31. Pumps. There are portable filter pumps listed by Underwriters Laboratories, and they comply with 680.31. They are both double insulated and equipped with an equipment grounding conductor that picks up the inaccessible dead metal parts, and they must have GFCI protection built in to their power supply cord.
680.32. Ground-Fault Circuit Interrupters Required. All 125-V receptacles (literally of any amperage, a clear error) within 6 m (20 ft) of the storable pool must have GFCI protection; the distances follow the same rules for measurement as in 680.22(A)(5) and do not pass through a sliding glass doorway, for example.
680.33. Luminaires. Underwater luminaires are available for storable pools in both 12-V versions with a transformer and in 120-V versions with long cords matching those that come with the motor. Usually these will not be subject to the default 450-mm (18-in.) submersion rule in 680.23(A)(5). Review the “through-wall lighting assembly” definition in 680.2 and the wiring rules for permanent pools in 680.23(E), and then factor in the following information from UL on products that are frequently installed improperly:
Underwater Luminaires for Aboveground Storable Swimming Pools—These luminaires are a type of through-wall lighting assembly as described in Art. 680 of the NEC. They have been investigated for use with an aboveground storable pool (a pool that is constructed on or above the ground and is capable of holding water to a maximum depth of 1.0 m (42 in.), or a pool with nonmetallic, molded polymeric walls regardless of dimension). They include all three of the following factory-provided parts:
1. Lamp assembly for temporary installation on or through the wall of an aboveground pool
2. Transformer or ground-fault circuit interrupter assembly provided with a 0.9–1.8 m (3–6 ft) power supply cord for connection to a source of supply and for temporary mounting away from the pool (the remote assembly)
3. Jacketed flexible cord of not less than 7.6 m (25 ft) in length connecting the lamp assembly and the remote assembly
These luminaires have been investigated for installation with the top of the lens not less than 200 mm (8 in.) below the top of the pool. A hole through the pool wall may be required for luminaire installation. Unless otherwise indicated in the luminaire’s installation instructions, the luminaire design has been investigated for the lower edge of any hole that a luminaire installer must cut in the pool wall to be no more than 360 mm (14 in.) below the top of the pool wall. The pool wall manufacturer may provide, at a greater depth, a properly sized hole or a reinforced wall section designed for field-cutting a properly sized hole for a luminaire or plumbing fitting. Unless otherwise marked for a maximum installation depth, these luminaires have been investigated for installation in such a hole at a greater depth where the pool installation instructions provide for the hole placement and usage.
Underwater Luminaires for Aboveground Nonstorable Swimming Pools—These luminaires are a type of through-wall lighting assembly as described in Art. 680 of the NEC. They have been investigated for permanent installation through or on the wall of an aboveground nonstorable pool. The information provided above for underwater luminaires for aboveground storable swimming pools regarding installation depth and using an existing hole or cutting a new hole for installation also applies to underwater luminaires for aboveground nonstorable swimming pools.
Convertible Underwater Luminaires for Aboveground Swimming Pools—These luminaires are initially configured as an underwater luminaire for aboveground storable swimming pool for use as described above. They include provisions for the one-time field conversion of the luminaire to an underwater luminaire for aboveground nonstorable swimming pool for use as described above. Once converted, these luminaires are not suitable for being modified back to their original configuration.
680.34. Receptacles. Receptacle outlets must be at least 1.8 m (6 ft) from the pool. How this can be enforced with what is in effect an appliance that can be set anywhere the ground is flat and for which no electrical permit will issue is anyone’s guess.
680.41. Emergency Switch for Spas and Hot Tubs. This switch is not a 680.12 maintenance disconnect. It is an emergency stop button for the spa motor, and may or (usually) may not operate across the line. It must be clearly labeled, adjacent to the unit and within sight, but at least 1.5 m (5 ft) away. The rule does not apply to single-family dwellings.
Its sole function is to allow one or more people using a spa, or even a passerby, usually in a commercial environment, to shut down the motor in the event someone becomes entrapped by suction from the intake of the pump. There have been a number of tragedies from this problem, with documented instances of children being eviscerated after sitting over an intake and the event in New Jersey that directly led to this requirement, where a woman was trapped and drowned as a number of football players tried and failed to pry her away from a broken intake. The real solution to this problem is nonelectrical and has been implemented in the product standards by now, in the form of multiple intakes widely separated so the complete obstruction of one only sends the suction elsewhere, changes to the intake port designs, and other safety improvements.
680.42. Outdoor Installations. Outdoor units must meet the rules in Part II governing permanently installed pools, except as they are varied in this section. There are four such changes. Part (A)(1) allows flexible connections in lengths up to 1.8 m (6 ft). This procedure is often used to bring a feeder up under a skirt of a spa or hot tub to a panel (makes the wiring a feeder) or to a “control panel” (the wiring lateral may be a branch circuit depending on conditions). Note that although this is written in terms of both liquidtight flexible metal and liquidtight flexible nonmetallic conduit, it only actually applies to the metallic version. The nonmetallic type is permitted in indefinite lengths in 680.25(A), and therefore the permission here is one that has already been granted. However, it is relevant for the metallic version. Note that this permission only applies to listed units, both the packaged equipment assembly versions and the self-contained spa or hot tub versions that include the tub vessel.
Part (A)(2) allows flexible cord up to 4.6 m (15 ft), thereby extending the usual 900 mm (3 ft) limit in 680.7. GFCI protection is required. Remember that the circuit ratings on this equipment typically run in the 40 A to 60 A range. Since this is an outdoor installation, 406.8(B)(2)(a) will ask for an in-use cover for this cord. This may necessitate a mobile-home type receptacle with an angle cord cap on the cord with its grounding pin set opposite to the cord exit.
Part (B) allows the omission of bonding connections to the metal bands or hoops that hold the wooden staves in place at a hot tub. All other requirements in 680.26 regarding equipotential bonding must be met.
Part (C) allows cabled wiring through the interior of a one-family dwelling or an accessory building thereto, such as a detached garage. Any Chap. 3 method is allowed as long as it includes a copper equipment grounding conductor of the requisite size that is incorporated within the cable assembly. This allows Type NM cable to run indoors to an outlet on the side of the house or outbuilding, or at that point a transition can also be made to other wiring methods that are recognized in 680.25(A) if the spa or hot tub location is some distance away.
Underwater luminaires, however, must follow the rules for underwater luminaires generally, which would involve a more restrictive list of wiring methods not including Type NM cable. Therefore they would need to be split from the feeder (because the last sentence requires compliance with 680.23, which presumes a qualified 680.25 feeder supplying the distribution. This is why the sentence was placed in this section to begin with; there was never any intent to make a variance for something as potentially hazardous as an underwater luminaire. The interior circuit would require a 680.23(F)(1) wiring method if the branch circuit were brought through, and if the feeder were left to supply all loads, it would need to follow 680.25(A) for its wiring methods. Another option given is a luminaire for a storable pool; this could be brought out on a conventional lighting circuit and the requisite luminaire, or luminaire and transformer, could be plugged in.
680.43. Indoor Installations. Indoor applications require Chap. 3 wiring with an allowance for a cord- and plug-connection for a listed unit rated 20 A or less. Part (A) requires that at least one 15- or 20-A, 125-V convenience receptacle must be installed at a spa or hot tub—not closer than 1.83 m (6 ft) from the inside wall of the unit and not more than 10 ft (3.0 m) away from it. This is intended to prevent the hazards of extension cords that might otherwise be used to operate radios, TVs, and so on (Fig. 680-20).
Fig. 680-20. At least one 15- or 20-A general-purpose receptacle must be installed at a spa or hot tub. (Sec. 680.43.)
As required by part (2) of this section, this receptacle and any others within 10 ft (3.05 m) of the spa or tub must be GFCI protected, including the one supplying the power to the unit. The distance measurements do not pass through doorways, etc., similar to other comparable provisions in the article.
Part (B)(1) classifies luminaires over a spa or hot tub by height, with the limitations becoming more severe as the luminaire descends. First, if the luminaire is at or over 3.7 m (12 ft) above (measured to maximum water level, which in this case would be the tub rim as per the definition, there is no further limitation. Second, if the lighting outlet is GFCI protected, the luminaire can be mounted as close as 2.3 m (7 ft 6 in.). Finally, if the luminaire is GFCI protected and recessed with either an electrically isolated metal trim or a nonmetallic trim, it can be even closer to the spa or hot tub. The same condition applies to surface-mounted luminaires with a glass or plastic globe and a nonmetallic body or a metallic body isolated from contact. In both of these circumstances the luminaire must be listed for a damp location.
Part (B)(2) governs underwater luminaires. Refer to the commentary on 680.42(C) at the end; the last sentence of that rule is the same as this wording and will be applied in the same way.
Part (C) requires the same 1.5-m (5-ft) exclusion zone around hot tubs and spas as 680.22(D) does for pools. There is no provision in this location that would recognize a barrier, so the distance cannot be decreased.
Part (D) requires bonding of the usual metal parts with 1.5 m (5 ft) of the spa or hot tub, including the pump motors. And there is the usual exception for incidental metal parts such as air and water jets, isolated plumbing fittings, towel bars, etc. Control devices within the 1.5 m (5 ft) zone must be bonded; if further away bonding is optional. A major change for the 2008 NEC is that by virtue of a new exception a listed self-contained spa or hot tub need not have its pump motors and the associated equipment meet the normal bonding rules. Now exactly what rules are being excused is unknown because the wording, in direct violation of the NEC Style Manual, does not form a complete sentence with a predicate. However, UL does know and we are likely to see changes in bonding conductor minimum sizes or other configuration changes under the skirt. Note that this exception does not reach listed packaged equipment assemblies where the tub vessel is provided separately.
Part (E) allows metal-to-metal contact on a frame or base and threaded piping interconnections for bonding methods, along with the usual 8 AWG solid copper. Part (F) requires grounding connections to the spa or hot tub equipment along with other electrical equipment within the 1.5 m (5 ft) zone, which would include such items as a thermostat for electric heat. Part (G) requires compliance with all rules in 680.27(A) for underwater audio equipment.
680.44. Protection. The default rule for all spas and hot tubs given here is that they will have GFCI protection in place for the outlet that supplies them. There are three circumstances where this will not be the case. The first concerns listed units, both self-contained units with the tub and equipment assemblies without, that have a system of integral GFCI protection for all components. Such units must be marked accordingly. The second exemption concerns field-assembled units that exceed in rating what is available in GFCI protective devices, specifically, over 250 V, or 3-phase, or over 50 A in heater load. Finally, where the spa or hot tub is part of a combination pool and spa or hot tub with a common bonding grid, it need not have the overall protection.
680.50. General. Part V of the article covers fountains, with a few exclusions. A fountain that is part of a body of water that in common with a pool is covered by the requirements for pools generally because they are more severe. Portable fountains are appliances and covered in Art. 422.
680.51. Luminaires, Submersible Pumps, and Other Submersible Equipment. All of this equipment must be GFCI protected unless listed for operation at 15 V or less and fed with a transformer that meets 680.23(A)(2) for equivalent pool equipment. Luminaires are limited in operating voltage to 150 V for the “supply circuits” and for submersible equipment including pumps to 300 V. The “supply circuit” wording allows for an HID luminaire with a higher starting voltage provided its line connection remained at 120 V. Note that the reach of this paragraph, that used to apply to all fountain equipment, now only applies to submersible equipment of the types specified. For example, there are fountains in outdoor northern areas that have 480-V heaters keeping them from freezing. This is not submersible equipment and beyond the reach of this paragraph.
Part (C) requires that luminaires must be installed with the lens below the normal water line, unless listed for use above water. If the luminaire faces up, the lens must be guarded against contact by the public, or it can be specifically listed for use without a guard. 680.23(A)(6) for regular swimming pools has a similar requirement, and UL requires distinctive installation instructions for those luminaires that can be used in the floor. Luminaires that are listed for use in fountains carry the term “submersible” in their description. For example, you might see a “wet-niche submersible luminaire,” or a “special use submersible luminaire.” These are not listed to the same requirements as luminaires listed for swimming pools, and they must not be confused with the other luminaires.
Part (D) states that luminaires and other equipment that must be submerged in order to avoid overheating must have a low-water cutoff or equivalent protection. In the case of luminaires, they will be marked “Submerse Before Lighting” and this will be visible after installation. This rule is similar to 680.23(A)(7) for lights in conventional swimming pools that must be submerged before lighting, which must be “inherently protected against the hazards of overheating.”
Part (E) limits equipment wiring entries such that they must either be threaded hubs or have suitable flexible cord, and any metal parts in contact with the water must be brass or other corrosion resistant material. Where flexible cord is used, it must not extend more than 3.0 m (10 ft) in the open within the fountain. If the cord extends beyond the fountain, it must run in an “approved wiring enclosure.” This could be a conduit, provided the conduit material was suitable for its location.
Part (F) requires equipment to be removable from the water for normal maintenance or relamping. Luminaires must not be embedded in the walls of the fountain such that the fountain must be drained in order to reach them for maintenance. Part (G) requires equipment in the fountain to be stable so it will not be likely to tip over, or it must be secured.
680.52. Junction Boxes and Other Enclosures. Junction boxes that aren’t located in the water must meet the same construction requirements as for enclosures connected to conduits leading to forming shells as covered in 680.24.
Part (B) covers underwater enclosures. These junction boxes and other submersible enclosures must:
Be made with threaded conduit entries, or else be provided with appropriate gland or other sealing connectors for flexible cord.
Be made of copper, brass, or other suitable corrosion-resistant material such as some types of stainless steel.
Be potted with an approved compound. Wax is often used for this purpose.
Be firmly attached to the fountain surface or supports, and bonded as required.
The remainder of the section applies 314.23(E) and (F). Nonmetallic conduit cannot serve as the sole support of a box, and if it is used, the enclosure must have additional supports that are suitably corrosion resistant so as to survive in the fountain water. If metal conduit is the support, two entries would be required, and it must have the requisite corrosion resistance.
680.53. Bonding. This rule requires piping to be bonded to the branch circuit equipment grounding conductor. An external bonding conductor wouldn’t be required, since the box must be connected to that conductor in its role as an equipment grounding conductor anyway, per 314.4.
The rule for bonding fountains is a very simple one, but it contrasts dramatically with the allowance in 680.26(A)(6) Exception that allows DI motors to be left out of the bonding grid. Metallic piping systems associated with a fountain must be bonded to the equipment grounding conductor of the branch circuit supplying the fountain. Note that this part of the article on fountains doesn’t specify how the bonding is to be achieved, or any minimum size bonding conductor, etc. It only states that certain piping is to be bonded.
680.54. Grounding. All electrical equipment within 1.5 m (5 ft) of the inside walls of the fountain, and all electric equipment associated with the recirculation system must be grounded. In addition, item (3) requires panelboards that supply a fountain to be grounded. Note that 408.40 imposes a grounding rule on all panelboards anyway.
680.55. Methods of Grounding. Essentially the rules that govern an underwater luminaire apply here. Specifically, the following rules for conventional pools, as listed in Part (A) come forward:
680.21(A), pool pump motor wiring
680.23(B)(3), equipment grounding provisions in cords to underwater luminaires
680.23(F)(1), wiring method rules for branch circuit wiring for underwater luminaires
680.23(F)(2), equipment grounding rules in the above branch circuits
680.24(F), equipment grounding provisions for supply circuits
680.25, feeder wiring methods and grounding
Part (B) has its own rule for grounding equipment supplied by cord and plug. The equipment grounding conductor must be an integral part of the cord, and it must connect the equipment grounding terminal at its supply end with all exposed non-current-carrying metal parts of the equipment. This is comparable to rules in 680.7, but it does not include the 900 mm (3 ft) limitation in 680.7(A) By setting the rule up this way, the cord can exceed 900 mm (3 ft) in length. This, in turn, correlates with the allowance for up to 3 m (10 ft) of flexible cord to run within the fountain, as covered in 680.51(E).
680.56. Cord-and-Plug-Connected Equipment. All electrical equipment under this heading, including the power-supply cords, must have GFCI protection ahead of it. Although this only applies to cord- and plug-connected equipment, 680.51(A) essentially requires GFCI protection on almost everything else.
Part (B) requires that flexible cord that is immersed in or subject to being splashed by the water must be rated for hard-service and also listed with a “W” suffix. This qualifies the cord for continuous submersion.
The ends of the flexible cord jacket, the conductor terminations, and the grounding terminations must all be set in a suitable potting compound so if water somehow gets into the cord, it cannot then get into the equipment. Now that listed potting compound is available for wet-niche grounding terminations [680.23(B)(2)(b)], this might be a good choice for these terminations as well.
Connections made with flexible cord for underwater equipment must be permanent. Although this paragraph allows for attachment plugs and receptacles, this is only for maintenance or storage of equipment that isn’t in an area of the fountain that contains water. In other words, submersible equipment supplied by a flexible cord cannot have that cord run out of the fixture to a receptacle in some remote, dry location.
680.57. Signs. This section covers signs in fountains, and also signs adjacent to fountains out to a distance of 3 m (10 ft) from the fountain. Part (B) requires GFCI protection on all circuits supplying a sign covered in this section. Part (C) prohibits fixed or stationary signs in fountains from being within 1.5 m (5 ft) from the inside walls of a fountain. Portable signs must not be placed within a fountain or outside a fountain within 1.5 m (5 ft) of the inside walls of the fountain. Part (D) requires local disconnects for the sign in accordance with both 680.12 and 600.6; both requirements can be met with a single device. Part (E) requires bonding connections in accordance with 600.7. Note that 600.7(B)(8) permits bonding in accordance with 680.53, which in turn enables a bonding connection to local metal piping.
680.58. GFCI Protection for Adjacent Receptacle Outlets. All 15- and 20-A 125-V through 250-V receptacle outlets within 6 m (25 ft) of a fountain must have GFCI protection. Note that this applies not just outdoors, but also within shopping malls and other indoor fountain locations. There is no allowance here for a waiver based on a locking configuration. This may prove problematic when running some maintenance equipment if the receptacle near the fountain is the only one for a great distance.
680.60. General. This is the beginning of Part VI on therapeutic tubs and pools. These are used for therapy in health care facilities, athletic training areas and similar areas. Portable therapeutic appliances need only comply with Art. 422 and don’t come under these requirements.
680.61. Permanently Installed Therapeutic Pools. If a therapeutic pool is installed in or on the ground, or within a building, such that it cannot be readily disassembled, then it must comply with all the rules for normal permanently installed pools, as covered in Parts I and II of the article. The only exception is for luminaires; they need not observe either the placement or the GFCI restrictions in 680.22(C)(1 through 4) if they are totally enclosed. The special rules in Part VI of the article discussed here only apply to therapeutic tubs that are essentially stationary. That is, they can be disassembled but are normally left in one place, per 680.62 (below).
680.62. Therapeutic Tanks (Hydrotherapeutic Tanks). These rules roughly correspond to 680.44 for (A), and 680.43 for (B) through (F) on indoor spas and hot tubs.
Part (A) requires that any of the subspecies of therapeutic tanks (See 680.2 for the following definitions: Packaged Therapeutic Tub or Hydrotherapeutic Tank Equipment Assembly; Self-Contained Therapeutic Tubs or Hydrotherapeutic Tanks, and then consider the undefined term used here, Field-Assembled Therapeutic Tub or Hydrotherapeutic Tank) must normally have GFCI protection. Listed self-contained units or listed packaged equipment assemblies that incorporate integral GFCI protection for electrical components need not be wired with additional GFCI protection. Field-assembled units with higher heating loads than 50 A, or that are three-phase or rated over 250 V don’t require GFCI protection.
Part (B) covers bonding. Just as swimming pools need a bonded environment for safety, tubs for therapeutic uses do as well. For stationary therapeutic tubs such as those at athletic training rooms and health care facilities (whether indoors or not), the following items need to be bonded together:
Any metallic fittings within or attached to the tub structure.
Metal parts of the circulation system, including the motor. Note that there is no allowance, at least as yet, for an unbonded DI pump motor.
Metal conduit, metal sheathed cables, and metal piping within 1.5 m (5 ft) of the inside walls of the tub, except items isolated by a permanent barrier. Note that this wording isn’t included in the bonding rules for conventional tubs in 680.43(D)(3). However, since 300.12 requires cable armors to be continuous between enclosures, and since the equipment must be bonded, presumably any metallic cable armor connected to a conventional spa or hot tub will be effectively bonded anyway.
All metal surfaces within 1.5 m (5 ft) and not separated by a permanent barrier from the tub area. This has the potential for including many small, incidental parts such as metallic rims on water jets and drain fittings (provided they aren’t connected to metallic piping), towel bars, mirror frames, etc. For this reason, 680.43(D)(4) Exception No. 1 covering conventional spas and hot tubs waives the bonding rule for small conductive surfaces such as these, that are not likely to become energized.
This exception has not been carried over into 680.62(B)(4) for therapeutic tubs. This appears to be an oversight. It is doubtful that those items are intended to have a bonding connection just because the tub is used for therapeutic purposes. The small parts exception corresponds to the allowance built into 680.26(B)(5), which waives the bonding rules at permanently installed pools for metal parts not over 100 mm (4 in.) in any dimension, and that don’t penetrate into the pool structure more than 25 mm (1 in.)
Electric devices and controls not associated with the tubs must be at least 5 ft away, unless they are bonded to the tub. This means that such items as a thermostat for electric heat must be bonded if they are within the 5 ft radius.
Note on the 1.5 m (5 ft) dimension. The Code is very unclear as to how this should be measured, in a radius or horizontally. The wording of 680.43(B)(1) implies a horizontal measurement by default, because luminaires are excluded from a 1.5 m (5 ft) zone to a height of 2.3 m (7½ ft) above the maximum water level. If we were to measure the 1.5 m (5-ft) distance in a radius, then a luminaire outside the tub would have to be 2.3 m (7½ ft) above the water outside the tub, but could be just over 1.5 m (5 ft) above the water level within the tub enclosure, which makes no sense. Nevertheless, the word “horizontal” is absent from this subsection, as well as subsection (D) on bonding, and the local authority must decide whether it should be inferred from the context. There isn’t any height limitation expressed either. The 3.7 m (12 ft) GFCI limitation in 680.43(B)(1)(a), and the same bonding limit in 680.26(B)(7) Exception No. 3 suggest that dimension as a reasonable limit. In addition, 680.43(A)(1) uses the phrase “measured horizontally” in the rule on receptacle placements.
Part (C) covers the methods of bonding. The bonding methods are similar for conventional and therapeutic tubs, but not identical. Both types of spas and tubs recognize the continuity of threaded metal piping entering a mated fitting. Both types also allow for metal-to-metal mounting on a common base or frame, just as for outdoor spas and hot tubs. Only therapeutic tubs, however, recognize “suitable metal clamps” for bonding. This may be a distinction without any practical difference, since metal-to-metal connections to a frame are recognized. Both types also require solid 8 AWG copper as the minimum for a separate bonding conductor. The solid conductor has far greater inherent resistance to chemical attack.
Part (D) covers grounding, and requires that all electrical equipment within 1.5 m (5 ft) of the inside walls of the facility, and that all electrical equipment associated with the recirculation system, must be connected to the equipment grounding conductor. Note that Chap. 6 rules cannot supersede rules in Chap. 5. If a therapeutic tub is used for patient care in a health care facility, then the enhanced grounding requirements in 517.13 will apply. An insulated copper equipment grounding conductor must be run to the unit, and the enclosure or armor of the wiring method must independently qualify as an equipment grounding return path. This includes metal-clad, armored, or mineral insulated cables with the outer armor fully qualified as a grounding path, most rigid metallic raceways, and some flexible metallic raceways up to 1.8 m (6 ft) in length.
Portable equipment must comply with 250.114 which requires equipment grounding connections for some equipment, and also the possibility of double-insulated appliances that will not have the grounding connection.
Part (E) requires receptacles within a 1.83 m (6 ft) radius to be GFCI protected. Part (F) requires luminaires to be totally enclosed, but does not impose and GFCI requirements on the lighting.
680.70. General. Unlike other general statements for parts within Art. 680, this one covering hydromassage bathtubs does not bring in other parts of the article. On the contrary, it positions this part as a stand-alone part, and no other part of Art. 680 applies to this equipment.
680.71. Protection. Any hydromassage bathtub and its associated electrical equipment must be supplied by an individual branch circuit and the outlet must be protected by a GFCI which must be readily accessible. All receptacles within 1.5 m (5 ft) of the tub and rated 125 V and not over 30 A must have GFCI protection as well. Note that most such tubs go in bathrooms where all the receptacles will be GFCI protected anyway, but these tubs do show up in other areas, and additional receptacles in the vicinity of the tub will require GFCI protection.
It is quite common to place the GFCI protection for one of these tubs as a GFCI receptacle adjacent to the pump motor, which is accessible (assuming 680.73 is met) but rarely readily accessible. The readily accessible requirement, new in the 2008 NEC, will change this practice. In theory, it should be possible in some locations to provide an access panel that is so well made, with obvious handles and unlatching mechanisms, set to open into a completely open area, and so uncluttered behind it, that an untrained person using no tools can quickly remove the panel and immediately find and put hands on the receptacle. That is possible, subject to the inspector’s judgment, but not likely. In reality, too many of these receptacle connections require a double-jointed midget with a mirror and a flashlight to find and reset. There are several good solutions to this. One is to use a GFCI circuit breaker, since the panelboard devices must be readily accessible. Another is to use a master-trip GFCI device, the so-called faceless GFCI that takes up one device gang and presents test and reset buttons but no receptacle slots. Remember that a remote GFCI feed through receptacle is not an option because, by virtue of its installation, the circuit it protects is no longer an individual branch circuit as this section requires.
680.72. Other Electric Equipment. Hydromassage bathtubs are not subject to the requirements for spas and hot tubs. Receptacles do not have to be at least 5 ft (1.5 m) from the tub’s inside wall, and luminaires do not have to be mounted at least 7½ ft (2.3 m) above the tub’s water level.
680.73. Accessibility. This rule reiterates the requirement of 110.26, but does so specifically directed to this equipment. Many hydromassage tub motors have been set behind tiled partitions for which there was no access whatsoever. Effective communication with the general contractor on this as well as many other issues is absolutely essential to the progress of work.
680.74. Bonding. All piping systems and all metal parts in contact with the circulation system must be bonded with a 8 AWG solid copper bonding conductor, connected to the bonding terminal on the motor. Double-insulated motors are exempt from this requirement. Note that the requirement is to bond piping systems. With today’s increasing use of nonmetallic water piping systems there is frequently nothing to bond under the skirt anyway. A metal escutcheon around the faucet with no metal piping behind it is not a metal piping system. It is very possible that even with a motor with a bonding lug, there will be no opportunity to run a bonding conductor. A bond wire must bond at least two things, and increasingly, there is no second item requiring bonding. Neither is it required to run a bonding conductor to the panel or anywhere else. Refer to the bonding discussion at 680.26 for the reasons why this is so.
682.1Scope. This article, new in the 2005 NEC, provides the first systematic coverage of electrical installations in or adjacent to bodies of water that do not qualify under Art. 680, such as artificial water features as part of landscaping, aeration ponds and storm retention basins, fish farms in artificial ponds, etc. As defined, the water depths can vary seasonally or be controllable. The article also addresses wiring in natural bodies of water, such as submersible pump in a natural pond for irrigation or other reasons.
682.2. Definitions. Aside from the descriptions of the bodies of water, three definitions are essential to apply this article. The most important is the electrical datum plane, which is (oversimplified) basically the horizontal plane that is 600 mm (2 ft) above the highest normal water level, such as the highest high tide. It is not the height of some catastrophic flooding event, such as the level of the 200-year flood as may have been established for other regulatory purposes such as flood insurance boundaries. The actual definition is in four parts and is much more complicated, but this will usually work.
An equipotential plane, for the purposes of this article is an area where wire mesh or other conductive elements are on, within, or under a walk surface not more than75 mm (3 in.) below the top surface, and these conductive elements are bonded to all metal structures and fixed nonelectrical equipment that may become energized, and connected to the electrical grounding system to prevent a difference in voltage from developing within the plane.
The shoreline is the farthest extent of standing water under the applicable conditions that determine the electrical datum plane for the specified body of water.
682.10. Electrical Equipment and Transformers. This is a good example of how the electrical datum plane is used in this article. Here electrical equipment and transformers including their enclosures must be specifically approved for the intended location. Only enclosures that are identified by their manufacturer’s published information as being suitable not just for a wet location, but rather for actual operation while submerged, can be used below the electrical datum plane, which is 2 ft above the normal high water mark.
682.11. Location of Service Equipment. This is a good example of where the shoreline definition is used. The service equipment for floating structures and submersible electrical equipment must be on land not closer than 1.5 m (5 ft) horizontally from the shoreline. The live parts must be elevated a minimum of 300 mm (12 in.) above the electrical datum plane, and there must be a shunt trip or equivalent arranged so the service disconnect will open if the water level reaches the height of the datum plane.
682.12. Electrical Connections. All electrical connections not intended for operation while submerged must be located at least 300 mm (12 in.) above the deck of a floating or fixed structure, but also, wherever located, not below the datum plane.
682.13. Wiring Methods and Installation. Liquidtight flexible metal conduit or liquidtight flexible nonmetallic conduit with approved fittings can be used for feeders and where flexible connections are required for services. Extra-hard usage portable power cable listed for both wet locations and sunlight resistance is also permitted for a feeder or a branch circuit where flexibility is required. Other wiring methods, suitable for the location can be installed where flexibility is not required. Temporary wiring that meets the provisions in 590.4 can also be used.
682.14. Disconnecting Means for Floating Structures and Submersible Electrical Equipment. Part (A) allows a circuit breaker, switch or both for this purpose, and requires that it be properly marked as to the load it disconnects, reiterating 110.22(A).
Part (B) covers the disconnect locations. It must be readily accessible, on land, and at least 1.5 m (5 ft) back from the shoreline. Its live parts must be elevated at least 300 mm (12 in.) above the datum plane. Note that this is a misuse of the defined term “live parts” and what is presumably intended is “uninsulated live parts”. An energized conductor is a live part regardless of whether it is insulated or not.
682.15. Ground-Fault Circuit Interrupter (GFCI) Protection. This covers 15-, 20-A 125- to 250-V receptacles that are outdoors, or if located in or on a floating building or a structure used for maintenance, storage, or repair. The rule applies to receptacles falling within the datum plane and that are used for portable electric hand tools, electrical diagnostic equipment, or portable lighting equipment. GFCI protection must be provided for these outlets, with the protective device located at least 300 mm (1 ft) above the datum plane.
682.30. Grounding. This section brings over rules from Arts. 553 (floating buildings), and 555 (marinas and boatyards), specifically 555.15. The 555.15 reference includes unremarkable, well known grounding rules, with one major exception, found in 555.15(B). Due to the corrosive effects of a marine environment, metallic raceways are not permitted to be relied on as equipment grounding conductors, and must, where used, be backed up by separate equipment grounding conductors, sized per Table 250.122 and [per 555.15(C)] not smaller than 12 AWG. The rules in Part III of Art. 553 contribute a requirement (from 553.11) that all metal parts in contact with water and any metal piping, as well as the usual metal parts that may become energized, must be connected to the equipment grounding system. No cross connections between neutral conductors and equipment grounding conductors are permitted, courtesy of 553.9.
682.31. Equipment Grounding Conductors. These rules are obvious and are taken from 555.15(C), 555.15(D), 555.15(E), and 553.10(B), as already incorporated by reference in the prior section.
682.32. Bonding of Non-Current-Carrying Metal Parts. This is taken from 553.11 (see above).
682.33. Equipotential Planes and Bonding of Equipotential Planes. This, start to finish, is critical information that breaks new ground. Much of this, although used in the NESC for utility work, is without precedent in the NEC. Review the definition in 682.2 before beginning. The section begins with the basic requirement to install such a plane as set forth in this section to mitigate “step and touch voltages” at electrical equipment. These terms are not defined in the NEC, but they are in the IEEE dictionary, as follows:
The step voltage is the potential difference between two points on the earth’s surface separated by a distance of one pace (assumed to be one meter) in the direction of maximum potential gradient. This potential difference could be dangerous when current flows through the earth or material upon which a worker is standing, particularly under fault conditions.
The touch voltage is the potential difference between a grounded metallic structure and a point on the earth’s surface separated by a distance equal to the normal maximum reach, approximately one meter. This potential difference could be dangerous and could result from induction or fault conditions, or both.
