A new NEC is on tap, with major changes in how the document presents the material, even changing to a new full-page format.

This analysis of the changes represented in the 1999 edition of the National Electrical Code (NEC) covers only the most significant of the many changes. Some of these changes are indeed major, including two new articles, (with another article deleted) and others are for minor editorial reasons. Major or minor, if you are involved with the particular subject matter the most minor change can assume major proportions in unexpected ways. Every proposal was made for a reason, and this magazine article is no substitute for careful study of the Code itself.

This series of four articles, running through the December issue, gives you an easy-to-find, in-depth analysis of the most important Code changes, with everything in Code order for ease of reference. This is, however, no substitute for a copy of the Code itself. Indeed, using this book without reference to the Code text under discussion would be counterproductive.

A good way to use the material is first review the changes and then discuss them with your colleagues until you are sure you understand how they work. Attend as many seminars as you can on this material. Make sure you can follow the technical discussion before you leave a topic. This time we're publishing two books that go into this material more comprehensively. Our "Electrician's Edition" includes all the material in the magazine articles, plus additional material we didn't have room to cover here, but equally well illustrated. Our "Engineer's Edition" includes all the material in the Electrician's Edition plus coverage of every single change in the NEC this cycle. Remember, Code changes, however minor, are only minor until they affect you; then they're major changes. These are the books that allow you to go deeper into all those other changes, especially in areas of the Code that affect you directly.

A few words about usability and exceptions. No general discussion of this Code cycle would be complete without mentioning the special NEC Usability Task Group. The group had enormous influence throughout the cycle with respect to major initiatives, such as the elimination of Art. 710 and the reorganization of Art. 250. You will be the judge of just how usable the Code has become, and we invite your letters and feedback at our seminars.

We cover those usability initiatives in sequence, except the elimination of exceptions by some panels. We have to discuss this here because it shows up from panel to panel throughout the Code. Refer to the sidebar, on page 34, for an example. The entire substantiation to make that change was: "Editorial. Convert 'Exceptions' into positive rules."

In the 1993 cycle, Sec. 240-21, for example, went from a simple rule with 11 exceptions (the tap rules) to the same set of tap rules written without exceptions. That well-received, entirely appropriate change became the unintended model in some panels to remove all exceptions in favor of positive text. As the sidebar shows, without very sophisticated editorial restructuring this very quickly results in successive, orphaned, unnumbered paragraphs that cannot be easily cited by a user or inspector, and that are in literal conflict with each other.

EC&M Tip No. 1: Read the entire cited section as a combined entity, then use the old principle of statutory interpretation; when two provisions conflict, read them so as to give maximum effect to both. If one paragraph says do it one way, and the next says in some cases you can do it another way, recognize the other possibility even though it would violate the first paragraph.

EC&M Tip No. 2: Be flexible, because you won't find editorial consistency. Not all Code making panels made these wholesale changes, and of those that did, some did a more sophisticated job than others. Art. 240 eliminated all of its over 25 former exceptions, but Art. 230 still has almost 40.

Art. 100, Receptacle. Once and for all, a receptacle is a single contact device no more. A duplex receptacle, for example, is two such contact devices located on the same yoke.

EC&M Tip: Never confuse a multiple receptacle with a single device.

Background: The prior Code did just that when it folded a fine print note into the rule, declaring it to be a single device "containing" two or more receptacles. That just didn't correlate with many other Code rules, such as Sec. 210-4(b). That section requires dwelling multiwire branch circuits to have a common disconnecting means if they serve "more than one device or equipment on the same yoke." When the definition changed in 1996, this section fell into question. How could you require the disconnect when the duplex receptacle was defined as a single device? Many other Code rules were also affected in unanticipated ways. The new definition resolves those correlation issues.

Art. 100, Service. Only an electric utility can be the source of a supply to a service. A dramatic and very controversial clarification in the definition, no doubt, but one that follows logically from revisions to the definition of "Service Conductors." Service conductors are now the conductors that extend from the service point to the service disconnecting means. The prior mention of other sources of electric power is no more.

EC&M Tip: You must know the policy of your local regulatory authorities regarding where the service point is: It varies from the street to the building, and everywhere in between. That point is the essential determinant of exactly how all connected conductors, on either side, get classified.

What happened: The long-standing definition of a separately derived system caught up with its logical implications. Other sources of on-site power are separately derived systems-pure and simple, and always have been. They are sources of power with no common connection with a supply on their line side. Now the Code is consistent, and the conductors derived from these sources are reclassified as feeders. Service equipment is redefined as well, and now referring only to service conductors; not supply conductors.

What about utility deregulation? Although utilities are being deregulated, you need to know what that really means. Electricity generation is being deregulated, but not distribution. The reason is simple; electric power distribution is what economists refer to as a "natural monopoly." Until someone figures out how to get several sets of power lines on a single set of utility poles, you won't have any choice as to who distributes that power. When competition is impossible, the result is regulation, and properly so.

The personnel maintaining the distribution system will be highly trained utility personnel, working in a regulated environment with access to specialized equipment. Until that changes in some fundamental way, the jurisdictional separation between codes based on the service point is defensible.

Sec. 110-10. This section says that you need to arrange the electrical system so a fault won't result in "extensive damage." Discussions often follow as installers and inspectors, manufacturers and testing agencies decide what that means. Two changes in this section should clarify the term. First, "current" ratings now replaces "withstand" ratings in the rule, and listed equipment installed in accordance with the listing is now presumed to comply.

What happened: The product standards carefully establish what the term "short-circuit current rating" is, and it differs from a "withstand rating." The Code, by changing the term referred to here now correlates the requirement with the ratings marked on electrical equipment. The final sentence, completely new, places the responsibility for determining equipment ratings on the testing laboratories and manufacturers. Once an inspector determines you installed listed equipment where it won't see fault current above its rating, then that ends the discussion.

Sec. 110-14(c)(1)d. Now motor terminations get a default 75C rating below 100A.

Background: This change corrects a problem with termination temperature ratings that just fell through the cracks. This section sets a normal temperature limit of 60C for small terminations, with an allowance for a higher listing. Normally that's how this problem would be fixed, but it won't work here. The reason is that motors seldom get listed, because testing laboratories don't know what the eventual shaft loading will turn out to be. That means the terminations can't qualify under the listing provisions. The NEMA motor standard provides for 75C conductors in its provisions, so there isn't a safety problem.

We always caution installers to remember conductors have two ends, and to consider the temperature rating of the load-side equipment as well as the supply circuit breaker or switch, etc. Now you can do just that with motors and not get into trouble, even where they aren't listed.

Sec. 110-22. The series-connected field labeling that used to have to show the available fault current at the point of the label, now only needs to show the general system available short-circuit current. This is much simpler, since the utility usually provides the available fault current at the service point.

EC&M tip: You still need to have an idea of what the available fault current is at these remote points. There are two reasons for this. First, in cases of minimal impedance, you're responsible to be sure the remote equipment is installed within its listing constraints. Second, always remember that this rule only applies where a series-connected listing is being relied upon to show adequate short-circuit duty rating. If you show that a feeder has enough impedance to reduce the available fault current below the interrupting rating of the overcurrent devices you're installing, then your devices aren't being used on a series-rated system, even if they would otherwise qualify. In this case, you don't need to field-mark a series application warning on the remote equipment.

Sec. 110-26(a)(3). Now you can set other electrical equipment ("associated with the electrical installation") out into the workspace required by this section, provided the intrusion isn't over 6 in.

Background: The prior Code clarified something that most electricians wished it hadn't. The dedicated workspace rule always prohibited any intrusion into the vertical plane defined by the front of the equipment. However, until the prior Code, you could avoid that conclusion. The 1996 Code, however, made that little fiction a thing of the past. This time the panel did a good job squarely facing the issue, giving us a specific distance limitation on the amount of intrusion. This should make for much more consistent expectations on the part of inspectors, and some needed flexibility in laying out electrical installations.

Sec. 110-26(f). The dedicated footprint rules for panelboards, switchboards, and motor control centers (formerly Sec. 384-4) has revised to reflect safety issues and relocated. The default dedicated zone comes down from 25 ft above the floor to 6 ft above the equipment, and no dedicated space applies below the equipment. You can make even further reductions in cases where you've provided the drip protection, and that's now something you'll need to be looking at regardless of height.

Note that Sec. 110-34(f) (formerly Sec. 710-9) says the same thing for over 600V.

Sec. 200-6. We'll be seeing a new type of identification on grounded conductors: three continuous white stripes.

The testing laboratories will decide just how wide these stripes need to be, or what percentage of the circumference of the conductor needs to be white. Hereafter, this section will be the only source for rules on grounded conductor identification. Similar provisions in Art. 210 and Art. 310 are deleted.

Sec. 200-7(c)(1). Now you can only reidentify a white wire, for use as an ungrounded conductor, when it's part of a cable assembly.

The panel decided that if you're dealing with a raceway, then there just isn't any good reason why you shouldn't pull in a conductor with properly colored insulation.

Sec. 200-7(c)(2). Now you have to reidentify the white wire in a cable assembly when you're using it as the supply run to the switch, as part of a switch leg. One of the oldest topics for beginning apprentices is how to run switch legs with cables. As long as the white wire went to the switch, then the black wire came back from the switch. This put the appropriate color contrast at the outlet. There would be one white wire loose from the supply circuit, and one black wire from the switch. This has been so ingrained in all of us for so long that the Code never asked us to reidentify that white wire going to the switch, even though it wasn't grounded. However, the panel reacted to an instance where an unqualified person didn't realize that white wire was hot.

Articles 210, 215, and 220 have been somewhat reorganized so Art. 220 only calculates loads, and it no longer tells us how a given load results in conductor selection and protection.

Here's a good example: Former Sec. 220-10(b) on continuous loads now appears in Sec. 215-2(a) for conductor ampacity and in Sec. 215-3 for overcurrent protective device sizing. Also, when a section covered two entirely different topics, the panel broke it up. For example, Sec. 210-52(d) is about mandatory receptacle placement rules in residential bathrooms, but in recent years it gained a 20A circuiting rule. Now that part of the rule joins similar material formerly in Sec. 220-4 as part of a new Sec. 210-11.

Sec. 210-11(c)(3) Ex. This new exception allows you to serve other equipment in a bathroom on a 20A bathroom receptacle circuit, provided it's the only bathroom served.

Background: This is a trade off between adding some load to a receptacle circuit in exchange for removing the possibility that other bathroom receptacles might be on the same circuit. Many shower ventilating fans come with a listing restriction calling for GFCI protection if installed in that area. The old Code required either a GFCI circuit breaker for these loads, or supply by a GFCI receptacle in an unusual location, such as the next bedroom, or if you had room in the box, a "master-trip" type GFCI without receptacle slots. In the face of those possibilities, the new exception looks pretty good.

Sec. 210-12. The Code defines, for the first time, the "arc-fault circuit-interrupter" (AFCI), which is a device that can recognize the electronic characteristics of an arcing fault in progress and then open the circuit. The AFCI concept makes its appearance, but the actual requirement only applies to dwelling unit bedroom circuits, and not until Jan. 1, 2002.

Background: The new devices are circuit breakers with electronic circuitry capable of recognizing the chaotic waveforms characteristic of electrical arcing. The devices are calibrated internally to trip, based on the likely ignition energy being released in the arc. The fault current levels, which produce secondary ignition of adjacent combustible materials, are reflected in the new UL standard.

As presently projected, an AFCI must clear a 5A arc in no more than 1 sec., and clear a 30A arc in no more than 0.11 sec. For higher level arcing faults, UL expects to use a test protocol that approximates someone cutting an energized cord with a pair of dikes, or a sharp edge of metallic furniture doing the same thing. The idea is to approximate a fault that isn't bolted but that does have a conductive surface in contact with the circuit conductors. In these cases, where the arc might be over 100A, the device must open with eight half cycles of fault initiation. The reason for the half-cycle parameter is because these arcs may sputter, and come and go. Therefore the arc may not last a full four cycles all at the same time, and yet it could do comparable damage by the end of the eighth half cycle.

One of the most difficult aspects of the new technology is coming up with a device that knows the difference between an arc from normal operations, such as unplugging a cord or opening a switch, and a destructive arc. In short, the devices cannot nuisance trip. The manufacturers involved with the proposals believe that they have solved this problem. There is even a device that will allow you to find hidden fault locations with a cheap AM radio.

EC&M Tips: Arc-fault detection does not replace other aspects of circuit protection. Short circuits from essentially bolted conditions must be still cleared instantaneously. Long term overloads must still be cleared on an inverse-time basis, with greater overloads clearing more rapidly. Although the arc-fault detection mechanism will be inside a typical molded-case circuit breaker, its sensing mechanism is strictly in addition to the conventional components. AFCI protection doesn't substitute for GFCI or GFPE protection, either. These devices are residual current devices: They measure the total current in a circuit, going to the load and returning from the load. If the vectorial summation of all currents does not equal zero, these devices detect the amount of that residual current and trip above a predetermined point.

AFCI protection does not duplicate any of these functions, although a manufacturer may well marry some of them within the same device. An AFCI detects and clears arcing faults whether line-to-line, or line-to-neutral, or line-to ground. In spite of outward similarities in appearance, don't confuse an AFCI circuit breaker and a GFCI circuit breaker.

Multiwire circuit functionality. Two-pole AFCI devices trip similarly to two-pole GFCI devices, by including the shared neutral in its sensing mechanism and then tripping as a two-pole device only. So, although you can protect a multiwire branch circuit with these devices, when one side trips, both circuits open. These devices don't have single-circuit functionality (that is, a two-pole device with independently-movable handles). Although this can't be done with a GFCI breaker, it could be done (with an additional sensor) with AFCI devices.

However, as this is written at least one major manufacturer has stated they have no intention to produce a two-pole device with multi-circuit functionality, and no other manufacturer has announced any plans to accommodate this feature. To the extent AFCI circuit (as opposed to outlet) protection, becomes mandatory, it may well mean the end of multiwire circuit functionality.

Sec. 210-52(c)(5), and Ex. Countertop receptacles need to go above, and no more than 18 in. above, the counter, except for construction designed for handicapped people. The exception for islands and peninsulas now only applies to flat countertops with no vertical breaks, such as backsplashes, etc., and with no cabinetry, etc. within 18 in. above, that you could put a receptacle in.

This result is a decent compromise between those who want no possibility of toddlers pulling hot appliances on top of them by yanking on cords, and those who routinely design these spaces. Relatively few consumers are willing to tolerate tombstone outlets in their kitchen designs. In the end there will be plenty of receptacles in the kitchen above wall-positioned counters that toddlers can't reach. The parent is the one to decide when the peninsula receptacle is safe to use, and not the National Electrical Code Committee. The changes also close a loophole in the prior Code, that arguably allowed any receptacle below a wall-positioned countertop to qualify as not being over 18 in. above that counter.

Sec. 210-52(d). The permitted location for the required bathroom basin receptacle has been nailed down. You have to put it on an adjacent wall, and not over 3 ft away from the outside edge of each basin.

The panel responded to tales of receptacles on a facing wall, but still close enough to be "adjacent" to the basin. The panel action asks whether there is a safety issue in effectively mandating two receptacles for two adjacent sinks. Although you can still put one receptacle outlet between the sinks, that's usually in the mirror and obviously problematic. Is anyone using personal care appliances on extension cords because the one receptacle outlet doesn't happen to fall next their preferred sink of the two? Both receptacles can, and ordinarily would go on the same branch circuit.

Sec. 210-52(g). The basement service receptacle needs to be in the service area. If part of the basement has been finished into a habitable room, then the receptacles in that room don't count as receptacle(s) required by this section. This clears up a long-standing area of controversy.

Sec. 210-60. In hotels and motels, only two guest room receptacles need to be readily accessible. In addition, if you do put a receptacle behind the bed, either put it low enough to be out of harm's way, or you need to suitably guard it. The prior Code attempted to require all the receptacles to be readily accessible, which was excessive. Many hotels intentionally design a receptacle for the credenza, for example, to be where it won't be readily accessible because it involves the electronic billing and cable television controls.

Guarding or relocating the receptacle behind the bed resolves a major area of damaged cords and receptacles. These problems are common and dangerous over time.

Sec. 210-70(a)(2). Outside lights now need only "provide illumination" for outdoor entrances and exits, which include remote floodlights. In addition, the lighting only needs to be provided for grade-level access points.

The prior wording used the preposition "at," which suggested the lighting outlet had to be adjacent to the doorway was never intended. In addition, entry lighting for such areas as a second floor deck with not grade level access is strictly a design issue.

Sec. 220-2(b). The fractions-of-an-ampere rule has been brought forward from its former location in the Chapter 9 Examples. If a calculation results in a fraction of 0.5A or higher, it has to be rounded up.

The examples are being moved to a new Appendix D, where they will officially lose all mandatory effect. This provision had to be rescued for obvious reasons.

EC&M tip: Don't make the mistake of rounding up twice in your calculations. The number "0.5" is a pure number. That means that if your calculation falls below this number, by whatever degree, it is below the number and can be dropped. Many people, confronted with a calculation that comes out xxx.499 (assume all figures to be significant), round the calculation to xxx.5. Then, they apply this rule and round that result to xxx + 1. Only round once, to the whole number of amperes, and at the conclusion of your calculations. Put differently, ask yourself, is this number large enough to be fully equal to one half, or is it smaller by any amount? If it is that large, and only if so, round it up. There are a numberof examination services that have this procedure incorrectly programmed in their answer keys.

Sec. 220-3(b)(9). A single piece of equipment that consists of four or more receptacles has to be figured at 90VA per receptacle.

Many manufacturers have been making this equipment for some time. Most inspection authorities, reasoning that since it occludes the area of two straps, it should count as at least 360VA. But, the Code never really addressed the issue. The drawing summarizes the rule and this new refinement. Note that this rule applies to calculations for receptacle outlets, which are features of the permanent wiring system. A plug-in adapter doesn't change an Art. 220 calculation for the wiring system.

Art. 225, Part B. All secondary building and structure disconnecting requirements have been relocated into this new part of the article, except medium voltage rules that go in a new Part C. The number of supply circuits to second or subsequent buildings or structures supplied from another is subject to limitation similar to the number-of-services limitations.

Sec. 225-30(d). Allowance included for multiple circuits as diverse sources of supply if used for "different uses" such as multiple control points for lighting. This avoids what could be routine, and harmless, violations of the principal rule. For example, a detached garage could have a power circuit to some receptacle outlets, and some three-way switch loops that may allow the owner to turn on the light at the outside of the garage door from the house, or to turn on the outside light at the house after getting out of the car in the garage.

Sec. 225-30(e). Additional supplies to multiple buildings are now also permitted for all occupancies with "documented safe switching procedures."

The concept of "safe switching procedures" has been around since the 1984 NEC, but only for large multibuilding industrial complexes, and never for additional supplies; only for remote switching. A documented safe switching procedure means whatever the AHJ says it does, and from jurisdiction to jurisdiction, this application could go anywhere from single-family homes with detached garages to large institutional campuses. This will take time to sort out and clarify, and you'll obviously want to discuss this one with the local authority.

Sec. 225-31. You need to supply a building or structure disconnecting means for circuits that merely pass through such a structure en route to supply another building or structure.

This is an important clarification. Energized conductors are energized conductors. They present equivalent risk whether or not they terminate within that particular building. However, the next section gives a way out.

Sec. 225-32. Now you can use the provisions of Sec. 230-6 to establish leeway in locating a building or structure disconnecting means location, even in cases where service entrance conductors aren't involved.

This is a very logical extension of the Sec. 230-6 concept, and it should be extremely useful.

Sec. 225-32 Ex. 1. The remote disconnecting means allowance, formerly only for large capacity industrial occupancies with "documented safe switching procedures" now extends to all occupancies with such procedures.

Refer to the discussion of Sec. 225-30(e). This change resulted from a defensible proposal to broaden the exception to include institutional occupancies. The idea was that many of these environments, including college campuses, have equally qualified electrical staff, and yet didn't qualify as industrial. The panel expanded this almost beyond recognition, however. This is another one you'll need to discuss with the local authority.

Sec. 230-9. The vertical clearance of final overhead spans needs to meet customary clearance requirements (as would be measured from grade) from elevated platforms from which they could be reached. This rule also applies within 3 ft, measured horizontally, of those areas.

This change resolves a controversy that stems from defining the limits of a porch, for example, whose physical manifestation only extends vertically to its railing. A conductor span terminating outside a 3-ft railing and 6 ft above the floor level complied with the old wording, because you would measure the clearance to grade and the distance from the porch structure would equal or exceed the 3-ft parameter. The new wording measures clearance from a 3-ft imaginary extension of the standing surface.

Sec. 230-40 Ex. 4. Now you can provide additional sets of service entrance conductors to supply common area loads as required in Sec. 210-25. Since these circuits can't be supplied from equipment serving one of the dwelling units, an additional set of service conductors are needed.

Sec. 230-40 Ex. 5. You can also have additional sets of service entrance conductors tapped from the line side of the disconnecting means in order to supply loads in Sec. 230-82, item 4. This resolves a correlation problem between these two sections.

Sec. 230-42(b). The minimum size rules for ungrounded service entrance conductors are deleted in favor of a reference to long-standing similar requirements in Sec. 230-79. This raises the minimum size of service conductors to a single-family house to 100A. It also lowers the size limit on single-circuit applications to No. 14 and for two-circuit applications to No. 10 (from No. 8).

Instead of having a parallel set of requirements as in the past, the panel went with a cross reference to Sec. 230-79, but the results are sometimes surprising, unsubstantiated, and perhaps unintended. For example, the minimum size on a single circuit was always No. 12. However, Sec. 230-79(a) sets the minimum disconnecting means for such circuits at 15A, which would allow a No. 14 for the first time. Similarly, the two-circuit installation used to be No. 8 minimum; now the rule goes over to the 30A disconnect size, which allows for a No. 10 instead. The former exception allowing a No. 8 for general use on limited demand situations by special permission has been deleted outright.

The one substantive change in Sec. 230-79 concerns single family homes, and its provisions would be reflected here as well. Now we can say good-bye to that journeyman's examination question, "when can a single-family home have a 60A service?" because the answer is never; the minimum size is 100A. The old rule that allowed 60A services if the initial load was below 10kVA with six or fewer branch circuits is a virtual dead letter. This is because with separate laundry and bathroom receptacle circuits, and with a minimum of two small-appliance branch circuits, any house with more than one lighting circuit couldn't use the allowance anyway.

EC&M tip: Don't jump to conclusions when you look at the exact phrasing of the rule calling for an ampacity not less than "the minimum rating of the disconnecting means." The word "minimum" is a theoretical minimum, and isn't dependent on standard sized device ratings. Suppose you have a calculated load of 237A. You still install 237A minimum conductors, subject to allowances elsewhere in the Code for continuous loads, etc. The fact that the switch might be 400A with 250A fuses, because those are standard sizes, doesn't mean you have to cable to the 400A switch size.

Sec. 230-46. Service entrance conductors can now generally be spliced with clamped or bolted connections. The connections need to be in an enclosure, or underground using listed underground splice kits.

This section used to state an outright prohibition followed by six exceptions with no end in sight. The panel decided there really wasn't any reason to micromanage the subject, and converted to a general allowance instead.

Sec. 230-50(a). You need to protect service entrance cables by raceway or other approved means if, and only if, they are subject to physical damage. The former list of areas that inspection authorities frequently took to be areas where liability to physical damage was to be presumed, has been deleted.

Sec. 240-3(d). This new subsection relocates the former obelisk notes to the ampacity tables into Art. 240.

The obelisk notes were overcurrent restrictions and not ampacity restrictions, so this is appropriate. In addition, a routine seminar topic always involves correlating the notes with such items as the motor circuit allowances in this section. Now that the material all sits in a common section, there should be less confusion.

Sec. 240-3(e). A tap conductor now gets an official definition as one with overcurrent protection ahead of its point of supply that exceeds the value for similar conductors protected as described elsewhere in Sec. 240-3.

A tap has historically been one of those things that everyone thinks they know what is, but turns out to have many very different meanings when you get into conversations with electricians. One interesting point is the definition doesn't mention transformer secondaries, which are recognized even within Art. 240, as tap sources.

The reason is that the tap section, Sec. 240-21, has been reorganized. Now all the transformer applications (with a transformer as the source) have their own subsection [Sec. 240-21(c)] and set of rules. Therefore we don't need to describe the conductor leaving the secondary as a tap anymore.

Sec. 240-4(b)(3). Now you can make an extension cord in the field from separately listed and installed components.

The old rule that only recognized listed extension cord sets has been modified. The wording includes listed components, not just cord bodies. The panel broadened the original proposal to include such items as an extension cord made with a listed 4-in. sq, box and double-duplex raised cover.

Sec. 240-21. The tap section has been restructured to split out conductors supplied by or through transformers into a different subsection; a new rule expressly forbids tapping of a tap.

Probably no single item causes more discussion at Code seminars than the seemingly endless variations on "Can I tap a tap." By long custom, the answer is no, but when you look at specific tap rules the literal text is far from clear. Inevitably seminar instructors end up talking about implications and intents. Now the arguments are over.

Sec. 240-33. The impracticability condition has been deleted from the horizontal enclosure rule for mounting overcurrent protective devices; now they need only comply with Sec. 240-81. Busway plug-in units, if listed, can set in any orientation consistent with the busway mounting position.

The rule still starts off with a requirement for vertical positioning, but follows with a sentence allowing horizontal positioning if consistent with Sec. 240-81. Sec. 240-81 says two things, namely, that circuit breakers have to be indicating (say whether "off" or "on") and that if they operate vertically (and not rotationally or horizontally), up must mean "on." If they mount horizontally, Sec. 240-81 is silent as to position. Since Sec. 240-33 is about circuit breakers mounted horizontally, any circuit breaker that is indicating will automatically comply.

This may not be the intended outcome. The proposal as submitted noted that the requirement originated in the 1926 Code, and retained the practicability restriction. The panel statement also pointed to listing restrictions as sufficient. However, the current UL guide card restrictions on molded-case circuit breakers say that circuit breakers may be mounted in any position unless marked otherwise.

Sec. 240-86(b). Series ratings can't be used on circuits where there is a motor contribution under fault conditions (figured as the total of the motor full-load currents) in excess of 1% of the interrupting rating of the lowest rated circuit breaker in the combination. This limitation only applies to motor load connected between the load side of the higher rated overcurrent device and the line side of the lower-rated overcurrent device.

A spinning motor becomes a generator under fault conditions, adding fault current. This rule reflects international practice and the ways series combinations are evaluated by testing laboratories. Although you need to account for it, motor contribution on the load side of the lower rated device is perfectly acceptable.

Sec. 240-90. A new Part H added to provide special allowances for supervised industrial installations "used exclusively for manufacturing or process control activities." Per a revision to the scope (Sec. 240-1), this part doesn't apply to medium voltage installations.

NOTE: The changes in this new part have unchallenged technical merit, particularly where huge transformers feed conductors difficult to install, staying within the typical 25-ft parameter. The principal challenge for our industry will be to provide the high degree of engineering expertise required, as well as sophisticated inspection to assure the intent of the rules is being observed.

Sec. 240-91. Supervised industrial installations defined as "industrial portions of a facility" with qualified maintenance and engineering supervision to assure qualified oversight. The Art. 220 load of just the process and manufacturing loads must be at least 2500kVA, and at least one service must be over 150V to ground and over 300V phase-to-phase.

EC&M tip: In determining both compliance with this limitation, and also the application of the subsequent provisions, be sure to exclude offices, warehouses, garages, machine shops, and recreational facilities. Keep two things in mind. First, these loads don't count in the allowable load to determine if you have a qualified industrial occupancy. Second, even if you have a qualified industrial occupancy, you can't serve these loads from the feeders supplied under these rules, unless, and only to the extent that this is true, these loads are an integral part of the industrial plant, substation, or control center. The 2500kVA limitation is a load limitation, and may be comprised of more than one service in its supply.

Sec. 240-92. This section includes the specific changes made for the supervised industrial locations, as follows:

  • Short-circuit and ground fault protection.

    You have three options. The primary device can serve, provided it isn't over 150% of the secondary conductor ampacity as reflected through the winding ratio, and the conductors aren't over 50 ft long. Another method allows 75-ft conductors with a differential relay arranged to trip no higher than the conductor ampacity. Finally, you can protect 75-ft conductors by other engineered methods that keep the time/current parameters within acceptable limits for all short-circuit and ground-fault conditions. One possibility might involve using fusible cable limiters. Although they don't provide overload protection, they have been used for years to isolate faulted cables, particularly in parallel runs.

  • Overload protection.

    Here again there are three options. You can terminate in a single overcurrent device, or you can terminate in up to six devices with a combined rating not over the ampacity of the conductors. The other option is to arrange overload current relaying and combining this with the ability to trip either the primary overcurrent device or enough downstream devices so the load current doesn't exceed the conductor ampacity. In addition to these three, the new rule recognizes that in some cases the short-circuit and ground-fault protective arrangements may provide overload protection as well. If engineering calculations show this to be the case, then separate overload protection isn't necessary.

  • Outside feeder taps.

    If the transformer is outdoors, then the secondary conductors can be of any length, as covered in Sec. 240-21(m). The only difference is that up to six overcurrent devices can be used at the feeder termination, instead of the single device permitted under normal circumstances.

Sec. 240-100. This is now a feeder and branch circuit combined section, with former Sec. 710-20 (discontinued) relocated as Sec. 240-100(a). The major substantive change occurs in the parent language in (a), which now requires the overcurrent device to be at the point of supply, or otherwise located under engineering supervision.

This breaks new ground, because there has never been a medium voltage overcurrent device location rule before. Sec. 240-101 now has the remainder of old Sec. 240-100, set as additional requirements for feeders.

Art. 250 has been completely reorganized with only one section (250-1) retaining its 1996 NEC identity.

Sec. 250-2. This new section combines the old Sec. 250-51 and the old Sec. 250-1 (FPN)s. In just one section we have a remarkably clear statement as to both the principles of grounding and the objectives of Art. 250. In effect, all of Art. 250 consists of prescriptive requirements that, when complied with, result in an electrical installation that delivers on the objectives set forth in this section.

The NEC is a prescriptive, not a performance-based, Code. That is, it doesn't tell us what electrical protective protocols are supposed to accomplish, it tells us instead what we are to do to accomplish the safety objectives in Sec. 90-1. The prescriptive requirements in the NEC result from a consensus process whereby the collective wisdom of international experience gets boiled down to actual requirements. If you want to know what a performance based Code would look like, imagine Art. 250 with just this one section. Everything else is essentially prescriptive.

Sec. 250-30(a)(1) [250-26(a)]. No parallel path allowed for grounded conductor return current over a main bonding jumper to the source of a separately derived system.

The main bonding jumper and the grounding electrode conductor connections to the grounded system conductor have to be made at the same point. There is an exception, but only for instances where there wouldn't be a parallel path, such as a nonmetallic conduit running back to a source so located as to not share conductive contact with common structural elements or common or connected equipment grounding conductors. The objective is to keep grounded conductor current confined to electrical conductors.

Sec. 250-30(a)(3), item 2 [250-26(c)(2)]. Water pipe electrodes for separately derived systems must follow the same distance-into-building constraints as similar electrodes for services.

This is a major change. On the one hand, remote water piping systems have been used for generations as grounding electrodes for separately derived systems. On the other hand, the entire reason for the into-the-building restriction was the likelihood of nonelectrical personnel disrupting the continuity of water piping systems; that outcome would be even more likely on a remote separately derived system. Now the latter aspect becomes the driving force for the new rule. EC&M tip: Don't get caught by this if you're a long way from the water entrance and there isn't any effectively grounded building steel nearby. Be sure you bid the job correctly if you run a grounding electrode conductor long distances. The exception for exposed piping in commercial buildings with qualified maintenance and supervision is retained, but be sure you can use it. The literal text refers to the "entire length" being exposed and you may need that conductor to pass through firestopped partitions. The panel intends such passages shouldn't kill the exception, but they also rejected a proposal allowing clarification. Make sure the local authority agrees.

Sec. 250-30(a)(3) Ex. [250-26]. A new exception allows manufacturers to ground a separately derived system originating in listed equipment used for service purposes to use the service grounding electrode and internal grounding bus, assuming the sizes are appropriate.

This is a common arrangement in unit substations.

Sec. 250-32(b)(2) [250-24(a)]. No parallel path is allowed for grounded conductor return current over a main bonding jumper to the source in an upstream building or structure. A separate equipment grounding conductor must be supplied and the grounded conductor must remain insulated from equipment grounding conductors in the second building.

There is an exception, but only for instances where there wouldn't be a parallel path, such as a nonmetallic conduit running back in an installation where no metal water pipes, etc., connect the two buildings. The objective is to keep grounded conductor current confined to electrical conductors.

Background: This stands on its head generations of Code writing, where the rule was to reground a neutral conductor at the second building, and only under an exception did you keep grounded and grounding conductors separate. Now that is reversed. The rule is to maintain separation, and a limited exception allows regrounding in some cases.

Sec. 250-50(a)(2) [250-81(a)]. Supplemental ground rods and other made electrodes of the pipe or plate type have to meet the resistance requirements of made electrodes generally, as covered in Sec. 250-56 [250-84].

This has the effect of requiring an additional rod or pipe electrode in those frequent instances in which the ground resistance of the first electrode exceeds 25 ohms.

Background: Some are suggesting since Sec. 250-56 allows for water pipes to augment a deficient made electrode, the water pipe plus the single electrode are enough. There are two problems with this. Basically, you're saying the electrode needing to be supplemented is capable of being a supplement, which is a needless repetition.

The second, far more important reason is that argument overlooks the reason for the rule in the first place. As long as the water pipe is in place, the ground rod is a mouse helping to carry the burden of an elephant. Its only function is to serve as the principal electrode if the metal water pipe is ever replaced with plastic. In that sense, a better term for it would be a reserve electrode. The intent of this change is absolutely clear: the made electrode system installed under this section must fully qualify as an electrode system with the metal water pipe removed.

Sec. 250-104(a)(4) [250-80(a)]. The local water piping system bonding requirement for separately derived systems now applies without regard to the type of grounding electrode employed.

The prior Code only imposed this requirement when the local electrode was a made electrode. That unraveled the intent of the requirement, which was to assure only minimal potential differences on nearby water piping systems, especially in view of the change in Sec. 250-30(a)(3), item 2. That change means that the actual grounding electrode connection for the separately derived system may be at a great distance from the local water piping.

Sec. 250-104(b) [250-80]. Aboveground portions of gas piping systems "upstream from the equipment shutoff valve" must be electrically continuous and bonded to the grounding electrode system.

This is extracted material from NFPA 54, the Fuel Gas Code. Although the NEC has always had a rule [formerly Sec. 250-80(b), now Sec. 250-104(c)] to bond other metal piping, this refers to piping that is connected to electrical equipment and that may become energized thereby. Normally, you comply with the rule when you connect the equipment grounding conductor, because you can use the equipment grounding conductor of the supply circuit for bonding. But, this applies even in cases where there isn't any electrical equipment connected to the gas piping system.

EC&M tip: The rule doesn't say how big this conductor should be. Because it applies to the gas piping system generally, the safest approach is to size in accordance with system bonding connections that also apply generally, such as to water piping systems. By that logic, size the conductor based on Table 250-66 [250-94]. You won't have to go back and do it again. Remember also, that a small size may be false economy. Sec. 250-64(a) [250-92(a)] as referenced in other subsections in Sec. 250-104 requires a raceway for sizes smaller than No. 6.

Sec. 250-122(f)(2) [250-95]. A new procedure allows for sizing equipment grounding conductors in paralleled cable assemblies based on the trip setting of upstream GFPE.

Now you can use Table 250-122 [250-95] to size equipment grounding conductors in parallel cable assemblies based on the trip setting of the GFPE, provided three things are true.

  • First, there must be qualified maintenance and supervision.

  • Second, the trip setting of the GFPE must not exceed the ampacity of a single ungrounded conductor in any oneof the cables run in parallel.

  • Third, the GFPE must be listed for this purpose. The panel was told that present GFPE equipment now available would probably prove suitable for this purpose and that the appropriate listing accommodations could be made fairly quickly.

Background: The normal rule for parallel circuits is that the full size equipment grounding conductor runs in every conduit. There are conditions under which one equipment grounding conductor will return almost all of the fault current, current that will back-feed into the fault on the other paralleled conductors through the common terminals. With raceway installations, this isn't a problem, but cable assemblies usually have an equipment grounding conductor appropriate to the likely overcurrent protection ahead of the ungrounded conductors.

This has meant that cable assemblies for parallel circuits usually need to be special ordered with oversized grounding conductors, which is a burden. The new change allows for conventional cable assemblies under conditions that assure that the circuit will open before any single grounding conductor would have to carry the monster fault assumed by the main rule.

Looking at the drawing, if the service GFPE is set to open at 300A, the cables as shown could be used. The cable ampacity is 310A, above the 300A trip setting, and a No. 4 equipment grounding conductor is appropriate for the same circuit size. If the exception were not used, then you would need to make up the cables with No. 1 equipment grounding conductors in each, corresponding to the 600A circuit protection.

Sec. 250-134 [250-57]. The equipment grounding rules of the Code apply to all grounded equipment, even if that grounding is by design and not by Code mandate.

Sec. 250-146(a) [250-74 Ex. 1]. Now you can use metal-to-metal contact as a receptacle equipment grounding path for recessed boxes, as well as surface ones.

Old work metal device boxes are held at the surface through the ears on the box. A receptacle screwed to such a box has the same surface area for metal-to-metal contact as a surface mounted handy box. Alternatively, some installers leave the ears on the receptacle. In this case, although the receptacle is a little far forward, the metal-to-metal surface area exceeds that present in a handy box.

EC&M tip: This rule depends on metal-to-metal contact. If your receptacles have fiber screw-retention washers, you must remove them in order to comply with this rule, whether you're using a flush box or a surface box.

Sec. 300-4(b)(1). The protective grommets for Type NM cable run through metal studs must cover the entire perimeter of the opening.

Years ago, inspectors used to come to meetings with pasteboard boxes full of Type NM cable shavings collected from construction sites where the cables had been abraded during installation on the unprotected upper edges of slots with V-shaped grommets. Although that isn't the problem it used to be, it is still a periodic issue. For example the cable box may hang while you're running the cable through the holes, and depending on the direction of tension, the cable may be stretched upward to where it runs over the upper edge of the slot. Sometimes a cable loop pulls out and catches on the unprotected upper edges as well.

This change addresses the problem. A little work with a pair of dikes can make two V-shaped grommets, one up and one down, cover an entire rectangular opening.

Sec. 300-11(a). Now you can put wiring methods back on suspended ceiling support wires, but only sometimes. The wires normally need to be in addition to those that are part of the ceiling design, they need to be secured at both ends, and in the case of a fire-rated ceiling assembly, you need to provide a means, such as color or tagging, so the inspector can tell which support wires you added.

Background: As in the prior code, you need to know the fundamental differences between a suspended ceiling that is an integral part of a fire-rated floor-ceiling or roof-ceiling assembly, which are comparatively unusual, and the far more common nonfire-rated assemblies. Some ceilings are designed with a thinner slab than one that would establish the required fire separation on its own. In this case, the suspended ceiling is an integral part of the overall fire rating. Qualified testing laboratories have evaluated many such designs, using specified ceiling panel materials. These tests involve actually constructing such a room and ceiling inside a giant furnace, so they are extremely expensive to run. Invariably, the designs specify, in precise detail, exactly how many support wires must be used, how thick they must be, how and where they must be attached, how fixtures are to be supported, how air ducts must be constructed and run, etc.

In making these tests, the testing laboratories do not assume any additional weight loads on these support wires, nor do they (or could they) evaluate how such a ceiling might deform under fire conditions with such loading. In these cases, the wiring methods must be supported securely and independently of the support wires. The only exception allows for the possibility someone might rerun a ceiling assembly test with specified wiring attached to specified supports. Because the rated designs specify the support wire locations, both electrical and building inspectors must be able to know which support wires are in addition to the essential elements of the ceiling construction.

In the case of nonfire-rated applications, you have to add wires to those required for ceiling support, but you don't have to identify them in any particular way. If the ceiling manufacturer chooses to recognize wiring on his specified ceiling support protocol, an exception allows you to use that recognition and avoid having to put up additional wires. Frankly the likelihood of that happening is only slightly better than infinitesimal, but the exception is probably harmless as long as no one expects to use it.

Sec. 300-14. The point that we measure from in determining the 6-in. minimum conductor length at boxes has been clarified as the point where the conductor emerges from its cable sheath or raceway in the box. To accommodate deep boxes, a new additional rule requires the free conductor length extend at least 3 in. beyond the front plane of the box, unless no dimension of the front opening is less than 8 in.

Background: This ends a controversy where some said the 6-in. rule applies from the front of the box and others said the back of the box. The difference is significant. If you have 12-2 NM cable with ground entering and then exiting a device box for a kitchen receptacle, that's 6 conductors in addition to the device. If the box is 4 in. deep, and the conductors have to be figured as yet another 6 in., the result would be a full 5 ft of No. 12 stuffed into that box. This rule also applies to equipment grounding conductors.

On the other hand, the 6-in. requirement won't allow us to work safely on conductors we can't reach. Anyone who's worn tools knows 6 in. doesn't make it when dealing with multiple extension rings or deep tile rings. The solution was to normally measure from the back of the box, and then be sure the wires would be long enough to work on. With large openings, however, if you can get both hands into the box to work on the conductors.

Sec. 300-21 (FPN). This fine print note now refers to the horizontal separation rules that "usually" apply, in recognition of recent listing recognition of systems that maintain fire separations without 24-in. horizontal separation of boxes on opposite sides of a wall.

One system uses special putty pads on the back of steel boxes. Another system uses special wall insulation and certain nonmetallic boxes. The normal requirement is still a 24-in. horizontal separation in order for the partition to retain its fire rating, even if studs intervene.

EC&M tip: You may find a limit on how many square inches of openings you can have per 100 square feet of wall surface, etc. When you see a nonmetallic box marked for use in a fire-rated assembly, that simply means you get to look up the assembly the box was tested for, and then apply the concomitant requirements. That marking does not mean that simply using the box preserves the indicated fire rating.

Sec. 305-6(a). The construction site GFCI requirements have been extended to 30A 125V receptacles. The cord set allowance for portable GFCI protection now includes devices that don't include cords, such as self-contained GFCI units.

These devices must be identified for portable use.

The requirement for identification as suitable for portable use means that the device has what most people refer to as open neutral protection. That is, if the grounded conductor opens for any reason, a relay drops out and the device deenergizes the load. Otherwise, an interrupted grounded conductor would cause the device to lose its brains and fail in its closed position.

Sec. 305-6(b). The assured equipment grounding conductor program (AEGCP) now clearly includes all other receptacle outlets used for the purposes in Sec. 305-6, of whatever voltage or amperage. Either GFCI or AEGCP applies, without exception.

This is very far reaching, although quite clear. One or the other protective protocol, GFCI or AEGCP, will apply to all receptacle outlets other than the 125V, 15A, or 20A, or 30A variety. Those outlets must have GFCI protection, although the allowance for qualified industrials to use AEGCP generally on these has been retained.

The other outlets get GFCI if made and if the facility wants it, otherwise AEGCP. This means that AEGCP would apply to a 200A pin-and-sleeve receptacle in a factory, if it were for Sec. 305-6 purposes, which are very broad.

EC&M tip: The AEGCP is a paperwork nightmare, so go with GFCI if you can. Remember also that AEGCP doesn't just apply to the activities of electrical personnel, far from it. It needs to be enforced on the general contractor affecting all trades, or on the person in charge of construction at a facility, because all the cords on site need to be documented.

That said, you may need to do just that more often than you thought. For example, there are residual current devices for 480Y/277V systems, but they aren't rated as GFCI protective devices because at this voltage, their entire trip curve characteristics would need to be reevaluated, perhaps with a faster trip.

Sec. 310-8(d). Conductors exposed to direct sunlight, including drip loops, must be listed for or marked as being sunlight resistant.

EC&M tip: Although everyone understands the value of this change for extended outdoor runs, such as single conductors in outdoor cable tray or cable bus, or open wiring on insulators, the lack of an exception for drip loops has the potential to create some very serious mischief. Although Type SE cable conductors go through some sunlight resistance testing, most building wire doesn't. It remains to be seen whether the major wire manufacturers are going to get their building wire reformulated and routinely tested for sunlight resistance. To the extent this is enforced, it may well raise hob with any raceway service entrance not located on the north side of a building.

Although many large black conductors would probably pass easily, for small services using No. 6 or smaller wire, the entire grounded conductor must be white. White wire can be made sunlight resistant, but it's expensive to do so. Meanwhile, no loss experience was ever cited in rejecting numerous public comments that sought an exception for drip loops. Be sure to sort this one out with the local authority, at least until you know what kind of conductor ratings will be locally available.

Sec. 310-15(b)(6). The term "main power feeder" has been defined for this application (the former Note 3) as the feeder(s) between the main disconnect and the lighting and appliance branch-circuit panelboard(s).

One of the problems in the prior Code concerned the fact that this term was undefined, and yet identified the feeder that qualified for the application of this special rule. If there were a 400A service, for example, that supplied two 200A lighting and appliance branch-circuit panelboards, which 200A feeder was the "main power feeder." The indications of possible multiple feeders at a dwelling unit ["feeder(s)"] leaves this still in doubt. The substantiation talked about only reaching the feeder that saw the normal diversity of the dwelling unit, but you'll need to sort out with your inspector how this will be applied.

Sec. 318-6(a). There are new provisions for discontinuous cable tray systems containing individual conductors instead of Chapter 3 wiring methods. The conductors can go from one tray to another, or from a tray to equipment, provided the transition distance isn't over 6 ft, the conductors are secured to the tray(s) at the transition, and the conductors are protected by either guarding or by location from physical damage.

Background: The Code continues to cover installations of Chapter 3 wiring methods in discontinuous trays. The wiring methods need only be secured in accordance with their applicable articles. In the case of Type MC cable, that would be 6 ft, and in the case of Type NM cable, 41/2 ft. However, there isn't any rule that covers single conductors. Although Sec. 320-5 generally sets a support interval for open wiring at 41/2 ft, Sec. 320-5(c) allows for a 30 ft spacing in industrial occupancies at 250 kcmil and larger. Since single conductors can only be used in cable trays when the tray runs in industrial occupancies, the frequent result would have otherwise been a 30 ft support interval, which no one supported. The result is a set of requirements that addresses the problem.

Sec. 336-5(a), item 1. Now you can use Type NM cable in single-family and two-family housing of any height or number of stories; however, multifamily and other uses remain subject to the 3-story limitation. Over the years there have been exceptions for basement conversions, attic conversions, etc. Now the three-story limitation simply disappears, along with those exceptions, for one-family and two-family dwellings.

Sec. 346-12(b)(3). The rule allowing a 20-ft free conduit drop in cases where no intermediate support is readily available has been broadened from "industrial machinery" to "stationary equipment or fixtures."

This change allows you to apply this outside of the industrial environment, and it has many practical applications in buildings with high ceilings. As before, the conduit must made up with threaded couplings, and it has to be firmly supported at the top and bottom of the riser. You can only use this procedure with rigid metal conduit.