per square foot from NEC Table 220.12. For residential applications, track lighting is considered to be included in the 33 VA/m² (3 VA/ft²) calculations and no additional load need be added.

If show-window lighting is supplied by a feeder, capacity must be included in the feeder to handle 200 VA per linear 300 mm (1 ft) of show-window length. Note that on the feeder level, the 200 VA per 300 mm (1 ft) is the only allowable calculation; unlike the branch circuit calculations that can be done either this way or on the basis of the receptacles actually in use taken at 180 VA each.

220.44. Receptacle Loads—Other Than Dwelling Units. This rule permits two possible approaches in determining the required feeder ampacity to supply receptacle loads in “other than dwelling units,” where a load of 180 VA of feeder capacity must be provided for all general-purpose 15- and 20-A receptacle outlets. (In dwelling units and in guest rooms of hotels and motels, no feeder capacity is required for 15- or 20-A general-purpose receptacle outlets. Such load is considered sufficiently covered by the load capacity provided for general lighting.) But in other than dwelling units, where a load of 180 VA of feeder capacity must be provided for all general-purpose 15- and 20-A receptacle outlets, a demand factor may be applied to the total calculated receptacle load as follows. Wording of this rule makes clear that either Table 220.42 or Table 220.44 may be used to apply demand factors to the total load of 180-VA receptacle loads when calculating required ampacity of a feeder supplying receptacle loads connected on branch circuits.

In other than dwelling units, the branch-circuit load for receptacle outlets, for which 180 VA was allowed per outlet, may be added to the general lighting load and may be reduced by the demand factors in Table 220.42. That is the basic rule of 220.44 and, in effect, requires any feeder to have capacity for the total number of receptacles it feeds and requires that capacity to be equal to 180 VA (per single or multiple receptacle) times the total number of receptacles (straps)—with a reduction from 100 percent of that value permitted only for the occupancies listed in Table 220.42.

Because the demand factor of Table 220.42 is shown as 100 percent for “All Other” types of occupancies, the basic rule of 220.44 as it appeared prior to the 1978 NE Code required a feeder to have ampacity for a load equal to 180 VA times the number of general-purpose receptacle outlets that the feeder supplied. That is no longer required. Recognizing that there is great diversity in use of receptacles in office buildings, stores, schools, and all the other occupancies that come under “All Others” in Tables 220.42, 220.44 contains a table to permit reduction of feeder capacity for receptacle loads on feeders. Those demand factors apply to any “nondwelling” occupancy.

The amount of feeder capacity for a typical case where a feeder, say, supplies panelboards that serve a total of 500 receptacles is shown in Fig. 220-7.

Although the calculation of Fig. 220-7 cannot always be taken as realistically related to usage of receptacles, it is realistic relief from the 100 percent-demand factor, which presumed that all receptacles were supplying 180-VA loads simultaneously.

220.50. Motors. Any feeder that supplies a motor load or a combination load (motors plus lighting and/or other electrical loads) must satisfy the indicated NEC sections of Art. 430. Feeder capacity for motor loads is usually taken at 125 percent of the full-load current rating of the largest motor supplied, plus the sum of the full-load currents of the other motors supplied. However, 430.26 allows the application of demand factors in certain cases as determined by the inspector.

Fig. 220-7. Table 220.13 permits demand factor in calculating feeder demand load for general-purpose receptacles. (Sec. 220.44.)

Specifically, 430.26 operates by “special permission,” which is the written permission from the authority enforcing the code. Some authorities recommend that no demand factor be used in determining the size of circuit to install so that the additional current capacity, thus allowed in the circuit, will give some spare capacity for growth. On the other hand, one of the major and repeated areas of discussion by the Code-making panel responsible for load calculations involves the repeated and well documented oversizing of industrial feeders in this area. Electric utilities know how much power they provide to their customers, and it too often does not compare well with load calculations run without appropriate demand factors. The express allowance for the judicious use of demand factors in the NEC for these loads is something that well deserves careful consideration. The factors given in Div. 12 of our sister book, the American Electricians’ Handbook, are an excellent place to start.

220.51. Fixed Electric Space Heating. Capacity required in a feeder to supply fixed electrical space-heating equipment is determined on the basis of a load equal to the total connected load of heaters on all branch circuits served from the feeder. Under conditions of intermittent operation or where all units cannot operate at the same time, permission may be granted for use of less than a 100 percent demand factor in sizing the feeder. 220.82, 220.83, and 220.84 permit alternate calculations of electric heat load for feeders or service-entrance conductors (which constitute a service feeder) in dwelling units. But reduction of the feeder capacity to less than 100 percent of connected load must be authorized by the local electrical inspector. Feeder load current for heating must not be less than the rating of the largest heating branch-circuit supplied.

220.52. Small-Appliance and Laundry Loads—Dwelling Unit. For a feeder or service conductors in a single-family dwelling, in an individual apartment of a multi-family dwelling with provisions for cooking by tenants, or in a hotel or motel suite with cooking facilities or a serving pantry, at least 1500 VA of load must be provided for each 2-wire, 20-A small-appliance circuit (to handle the small appliance load in kitchen, pantry, and dining areas) that is actually installed. The total small-appliance load determined in this way may be added to the general lighting load, and the resulting total load may be reduced by the demand factors given in Table 220.42.

Note that in a major clarification, the 2008 NEC changed the verb in this rule from “required by 210.11(C)(1)” to “covered by 210.11(C)(1).” A key point of contention for a very long time has been whether 3000 VA based on two small appliance branch circuits was all that were needed to put into a load calculation, even if many more were actually installed in a given dwelling unit. The theory behind this was that 210.11(C)(1) only requires two such circuits. The opposing viewpoint noted that 210.11(C)(1) actually mandated “two or more” such circuits, so all that were provided should be counted in the load calculation. By changing the word, it is now clear that although only two such circuits are required, if you choose to install more, the load calculation must include every one that is in place, taken at 1500 VA each. The same change was made in (B) following, so if multiple laundry circuits are provided, each will enter the load calculation at 1500 VA each.

A feeder load of at least 1500 VA must be added for each 2-wire, 20-A laundry circuit installed as covered by 210.1(C)(2). And that load may also be added to the general lighting load and subjected to the demand factors in Table 220.42.

220.53. Appliance Load—Dwelling Unit(s). For fixed appliances (fastened in place) other than ranges, clothes dryers, air-conditioning equipment, and space-heating equipment, feeder capacity in dwelling occupancies must be provided for the sum of these loads; but, if there are at least four such fixed appliances, the total load of four or more such appliances may be reduced by a demand factor of 75 percent (NE Code 220.53). Wording of this rule makes clear that a “fixed appliance” is one that is “fastened in place.”

As an example of application of this Code provision, consider the following calculation of feeder capacity for fixed appliances in a single-family house. The calculation is made to determine how much capacity must be provided in the service-entrance conductors (the service feeder):

To comply with 430.24, 25 percent is added to the full-load current of the ½-hp, 120-V appliance motor because it is the highest-rated motor in the group. Since it is assumed that the load on the 120/240-V feeder will be balanced and each of the ¼-hp motors will be connected to different ungrounded conductors, only one is counted in the preceding calculation. Except for the 120-V motors, all the other appliance loads are connected to both ungrounded conductors and are automatically balanced. Since there are four or more fixed appliances in addition to a range, clothes dryer, etc., a demand factor of 75 percent may be applied to the total load of these appliances. Seventy-five percent of 33.4 = 25 A, which is the current to be added to that computed for the lighting and other loads to determine the total current to be carried by the ungrounded (outside) service-entrance conductors.

The preceding demand factor may be applied to similar loads in two-family or multifamily dwellings.

220.54. Electric Clothes Dryers—Dwelling Unit(s). This rule prescribes a minimum demand of 5 kVA for 120/240-V electric clothes dryers in determining branch-circuit and feeder sizes. Note that this rule applies only to “household” electric clothes dryers, and not to commercial applications. This rule is helpful because the ratings of electric clothes dryers are not usually known in the planning stages when feeder calculations must be determined (Fig. 220-8).

Fig. 220-8. Feeder load of 5 kVA per dryer must be provided if actual load is not known. (Sec. 220.54.)

When sizing a feeder for one or more electric clothes dryers, a load of 5000 VA or the nameplate rating, whichever is larger, shall be included for each dryer—subject to the demand factors of Table 220.54 when the feeder supplies a number of clothes dryers, as in an apartment house. At one time this table periodically generated paradoxical load calculations; for some load brackets, adding additional clothes dryers actually decreased the calculated load for the feeder. This has been corrected, and now adding a clothes dryer always results in at least some additional load capacity required in the feeder.

220.55. Electric Ranges and Other Cooking Appliances—Dwelling Unit(s). Feeder capacity must be allowed for household electric cooking appliances rated over 1¾ kW, in accordance with Table 220.55 of the Code. Feeder demand loads for a number of cooking appliances on a feeder may be obtained from Table 220.55.

Note 4 to Table 220.55 permits sizing of a branch circuit to supply a single electric range, a wall-mounted oven, or a counter-mounted cooking unit in accordance with that table. That table is also used in sizing a feeder (or service conductors) that supplies one or more electric ranges or cooking units. Note that 220.55 and Table 220.55 apply only to such cooking appliances in a “dwelling unit” and do not cover commercial or institutional applications, although ranges in vocational school kitchens are covered.

Figure 220-9 shows a typical NEC calculation of the minimum demand load to be used in sizing the branch circuit to the range. The same value of demand load is also used in sizing a feeder (or service conductors) from which the range circuit is fed. Calculation is as follows:

A branch circuit for the 12-kW range is selected in accordance with Note 4 of Table 220.55, which says that the branch-circuit load for a range may be selected from the table itself. Under the heading “Number of Appliances,” read across from “1.” The maximum demand to be used in sizing the range circuit for a 12-kW range is shown under the heading “Maximum Demand” to be not less than 8 kW. The minimum rating of the range-circuit ungrounded conductors will be

Fig. 220-9. Minimum amp rating of branch-circuit conductors for a 12-kW range. (Sec. 220.55.)

NE Code Table 310.16 shows that the minimum size of copper conductors that may be used is 8 AWG (TW—40 A, THW—45 A, XHHW or THHN—50 A). No. 8 AWG is also designated in 210.19(A)(3) as the minimum size of conductor for any range rated 8¾ kW or more because the circuit must be at least rated 40 A.

The overload protection for this circuit of No. 8 TW conductors would be 40-A fuses or a 40-A circuit breaker. If THW, THHN, or XHHW wires are used for the circuit, they must be taken as having an ampacity of not more than 40 A and protected at that value. That requirement follows from the UL rule that conductors up to No. 1 AWG size must be used at the 60°C ampacity for the size of conductor, regardless of the actual temperature rating of the insulation—which may be 75 or 90°C. Similarly, 110.14(C)(1)(a) brings the same listing limitations into the NEC itself. The ampacity used must be that of TW wire of the given size.

Although the two hot legs of the 120/240-V, 3-wire circuit must be not smaller than No. 8, Exception No. 2 to Sec. 210.19(A)(3) permits the neutral conductor to be smaller, but it specifies that it must have an ampacity not less than 70 percent of the rating of the branch-circuit CB or fuse and may never be smaller than No. 10.

For the range circuit in this example, the neutral may be rated

70% × 40 A (rating of branch-circuit protection) = 28 A

This calls for a No. 10 neutral.

Figure 220-10 shows a more involved calculation for a range rated over 12 kW. Figure 220-11 shows two units that total 12 kW and are taken at a demand load of 8 kW, as if they were a single range. Figure 220-12 shows another calculation for separate cooking units on one circuit. And a feeder that would be used to supply any of the cooking installations shown in Figs. 220-9 through 220-12 would have to include capacity equal to the demand load used in sizing the branch circuit.

Fig. 220-10. Sizing a branch circuit for a household range over 12 kW. (Sec. 220.55.)

Fig. 220-11. Two units treated as a single-range load. (Sec. 220.55.)

A feeder supplying more than one range (rated not over 12 kW) must have ampacity sufficient for the maximum demand load given in Table 220.55 for the number of ranges fed. For instance, a feeder to 10 such ranges would have to have ampacity for a load of 25 kW.

Other Calculations on Electric Cooking Appliances

The following “roundup” points out step-by-step methods of wiring the various types of household electric cooking equipment (ranges, counter-mounted cooking units, and wall-mounted ovens) according to the NEC.

Tap Conductors

210.19(A)(3), Exception No. 2, gives permission to reduce the size of the neutral conductor of a 3-wire range branch circuit to 70 percent of the rating of the CB or fuses protecting the branch circuit. However, this rule does not apply to smaller taps connected to a 50-A circuit—where the smaller taps (none less than 20-A ratings) must all be the same size. Further, it does not apply when individual branch circuits supply each wall- or counter-mounted cooking unit and all circuit conductors are of the same size and less than No. 10.

Fig. 220-12. Determining branch-circuit load for separate cooking appliances on a single circuit. (Sec. 220.55.)

210.19(A)(3), Exception No. 1, permits tap conductors, rated not less than 20 A, to be connected to 50-A branch circuits that supply ranges, wall-mounted ovens, and counter-mounted cooking units. These taps cannot be any longer than necessary for servicing. Figure 220-13 illustrates the application of this rule.

Fig. 220-13. One branch circuit to cooking units. (Sec. 220.55.)

In 210.19(A)(3), Exception No. 1, the wording “no longer than necessary for servicing” encourages the location of circuit junction boxes as close as possible to each cooking and oven unit connected to 50-A circuits. A number of counter-mounted cooking units have integral supply leads about 36 in. (914 mm) long, and some ovens come with supply conduit and wire in lengths of 48 to 54 in. Therefore, a box should be installed close enough to connect these leads.

Feeder and Circuit Calculations

220.55 permits the use of Table 220.55 for calculating the feeder load for ranges and other cooking appliances that are individually rated more than 1¾ kW.

Note 4 of the table reads: “The branch-circuit load for one wall-mounted oven or one counter-mounted cooking unit shall be the nameplate rating of the appliance.” Figure 220-14 shows a separate branch circuit to each cooking unit, as permitted.

Common sense dictates that there is no difference in demand factor between a single range of 12 kW and a wall-mounted oven and surface-mounted cooking unit totaling 12 kW. This is explained in the last sentence of Note 4 of Table 220.55. The mere division of a complete range into two or more units does not change the demand factor. Therefore, the most direct and accurate

Fig. 220-14. Separate branch circuit to cooking units. (Sec. 220.55.)

method of computing the branch-circuit and feeder calculations for wall-mounted ovens and surface-mounted cooking units within each occupancy is to total the kilowatt ratings of these appliances and treat this total kilowatt rating as a single range of the same rating. For example, a particular dwelling has an 8-kW, 4-burner, surface-mounted cooking unit and a 4-kW wall-mounted oven. This is a total of 12 kW, and the maximum permissible demand given in Column C of Table 220.55 for a single 12-kW range is 8 kW.

Similarly, it follows that if the ratings of a 2-burner, counter-mounted cooking unit and a wall-mounted oven are each 3.5 kW, the total of the two would be 7 kW—the same total as a small 7-kW range. Because the 7-kW load is less than 8¾ kW, Note 3 of Table 220.55 permits Column B of Table 220.55 to be used in lieu of Column C. The demand load is 5.6 kW (7 kW times 0.80). Range or total cooking and oven unit ratings less than 8¾ kW are more likely to be found in small apartment units of multifamily dwellings than in single-family dwellings.

Because the demand loads in Column C of Table 220.55 apply to ranges not exceeding 12 kW, they also apply to wall-mounted ovens and counter-mounted cooking units within each individual occupancy by totaling their aggregate nameplate kilowatt ratings. Then if the total rating exceeds 12 kW, Note 1 to the table should be used as if the units were a single range of equal rating. For example, assume that the total rating of a counter-mounted cooking unit and two wall-mounted ovens is 16 kW in a dwelling unit. The maximum demand for a single 12-kW range is given as 8 kW in Column C. Note 1 requires that the maximum demand in Column C be increased 5 percent for each additional kilowatt or major fraction thereof that exceeds 12 kW. In this case 16 kW exceeds 12 kW by 4 kW. Therefore, 5 percent times 4 equals 20 percent, and 20 percent of 8 kW is 1.6 kW. The maximum feeder and branch-circuit demand is then 9.6 kW (8 kW plus 1.6 kW). A 9600-W load would draw exactly 40 A at 240 V, thereby just fitting on a circuit rated 40 A.

For the range or cooking unit demand factors in a multifamily dwelling, say a 12-unit apartment building, a specific calculation must be made, as follows:

1. Each apartment has a 6-kW counter-mounted cooking unit and a 4-kW wall-mounted oven. And each apartment is served by a separate feeder from a main switchboard. The maximum cooking demand in each apartment feeder should be computed in the same manner as previously described for single-family dwellings. Since the total rating of cooking and oven units in each apartment is 10 kW (6 kW plus 4 kW), Column C of Table 220.55 for one appliance should apply. Thus, the maximum cooking demand load on each feeder is 8 kW.

2. In figuring the size of the main service feeder, Column C should be used for 12 appliances. Thus, the demand would be 27 kW.

As an alternate calculation, assume that each of the 12 apartments has a 4-kW counter-mounted cooking unit and a 4-kW wall-mounted oven. This would total 8 kW per apartment. In this case Column B of Table 220.55 can be used to determine the cooking load in each separate feeder. By applying Column B on the basis of a single 8-kW range, the maximum demand is 6.4 kW (8 kW times 0.80). Therefore, 6.4 kW is the cooking load to be included in the calculation of each feeder. Notice that this is 1.6 kW less than the previous example where cooking and oven units, totaling 10 kW, had a demand load of 8 kW. And this is logical, because smaller units should produce a smaller total kilowatt demand.

On the other hand, it is advantageous to use Column C instead of Column B for computing the main service feeder capacity for twelve 8-kW cooking loads. The reason for this is that Column B gives a higher result where more than five 8-kW ranges (or combinations) and more than twelve 7-kW ranges (or combinations) are to be used. In these instances, calculations made on the basis of Column B result in a demand load greater than that of Column C for the same number of ranges. As an example, twelve 8-kW ranges have a demand load of 30.72 kW (12 times 8 kW times 0.32) in applying Column B, but only a demand load of 27 kW in Column C. And in Column C the 27 kW is based on twelve 12-kW ranges. This discrepancy dictates use of Column B only on the limited basis previously outlined.

The reason for the higher demand factor for a smaller range is that the smaller appliances, while in use, have more of their current consuming functions in operation at the same time. This is the reason that the demand factors in Column A are never lower and usually higher for the same number of appliances listed in Column B. When the cooking function is subdivided over several appliances, as in this example, the effect is to create one large range, and collectively the appliances will behave as a larger range with its attendant effectively smaller demand factor in Column C. Note 3 gives the option of using either Column C or Columns A and/or B to make this calculation for this reason.

Branch-Circuit Wiring

Where individual branch circuits supply each counter-mounted cooking unit and wall-mounted oven, there appears to be no particular problem. Figure 220-14 gives the details for wiring units on individual branch circuits.

Figure 220-13 shows an example of how typical counter-mounted cooking units and wall-mounted ovens are connected to a 50-A branch circuit.

Several manufacturers of cooking units provide an attached flexible metal conduit with supply leads and a floating 4-in. octagon box as a part of each unit. These units are commonly called “prewired types.” With this arrangement, an electrician does not have to make any supply connections in the appliance. Where such units are connected to a 50-A circuit, the 4-in. octagon box is removed, and the flexible conduit is connected to a larger circuit junction box, which contains the 6 AWG circuit conductors.

On the other hand, some manufacturers do not furnish supply leads with their cooking units. As a result, the electrical contractor must supply the tap conductors to these units from the 50-A circuit junction box (see Fig. 220-13). In this case, connections must be made in the appliance as well as in the junction box. Figure 220-15 shows a single branch circuit supplying the same units, as shown in Fig. 220-13.

Fig. 220-15. Separate circuits have advantages. (Sec. 220.55.)

40-A Circuits

The NEC does recognize a 40-A circuit for two or more outlets, as noted in 210.23(C). Because an 8 AWG (40-A) circuit can supply a single range rated not over 16.4 kW, it can also supply counter- and wall-mounted units not exceeding the same total of 16.4 kW. The rating of 16.4 kW is determined as the maximum rating of equipment that may be supplied by a 40-A branch circuit, which has a capacity of 9600 W (40 A × 240 V). From Note 1 to Table 220.55, a 15.4-kW load would require a demand capacity equal to 8000 W plus [(16.4 × 12) × 0.05 × 8000] = 8000 W plus 4 × 0.05 × 8000 = 8000 plus 1600 = 9600 W.

Figure 220-16 shows an arrangement of a 8-AWG (40-A) branch circuit supplying one 7.5-kW cooking unit and one 4-kW oven. Or individual branch circuits may be run to the units.

220.56. Kitchen Equipment—Other than Dwelling Unit(s). Commercial electric cooking loads must comply with 220.56 and its table of feeder demand factors for commercial electric cooking equipment—including dishwasher booster heaters, water heaters, and other kitchen equipment. Space-heating, ventilating, and/or air-conditioning equipment is excluded from the phrase “other kitchen equipment.”

At one time, the Code did not recognize demand factors for such equipment. Code Table 220.56 is the result of extensive research on the part of electric utilities. The demand factors given in Table 220.56 may be applied to all equipment (except the excluded heating, ventilating, and air-conditioning loads) that is either thermostatically controlled or is used only on an intermittent basis. Continuously operating loads, such as infrared heat lamps used for food warming, would be taken at 100 percent demand and not counted in the “Number of Units” that are subject to the demand factors of Table 220.56.

The rule says that the minimum load to be used in sizing a feeder to commercial kitchen equipment must not be less than the sum of the largest two kitchen equipment loads. If the feeder load determined by using Table 220.56 on the total number of appliances that are controlled or intermittent and then adding the sum of load ratings of continuous loads like heat lamps is less than the sum of load ratings of the two largest load units—then the minimum feeder load must be taken as the sum of the two largest load units.

example Find the minimum demand load to be used in sizing a feeder supplying a 20-kW quick-recovery water heater, a 5-kW fryer, a dough mixer with a 3-phase 1½-hp, 208-V motor, and four continuously operating 250-W food-warmer infrared lamps—with a 208Y/120V, 3-phase, 4-wire supply.

Although the water heater, the fryer, and the four lamps are a total of 1 + 1 + 4, or 6, unit loads, the 250-W lamps may not be counted in using Table 220.56 because they are continuous loads. For the water heater and the fryer, Table 220-56 indicates that a 90 percent demand must be used where the “Number of Units of Equipment” is 3. The motor must be taken at 125 percent per 430.24, and based on Table 430.250 per 430.6(A)(1). The table current is 6.6 A, and 125 percent of that is 8.3 A.

To convert the motor load to volt-amperes, do the following multiplication: 8.3 A × 208 V ×. Although this should be familiar to you, and it is the usual

Fig. 220-16. A single 40-A circuit may supply units. (Sec. 220.55.)

1.732 term used in three-phase work all the time, remember that 208 V is simply the line-to-neutral voltage, also multiplied by the same term, thus the same multiplication can be written:

8.3 A × 120 V × × . Since × = 3, the multiplication is simply 8.3 A × 360 V. Working with 360 in 208Y/120 V systems (and 831 in 480Y/277 V systems) is much simple and faster than using the square root of three and the line-to-line voltage.

Therefore, 8.3 A × 360 V = 3 kVA, and feeder minimum load (kW = kVA) must then be taken as

Four 250-W lamps @ 100% +1.0 kVA = 26.2 kVA

Then, the feeder must be sized for a minimum current load of

The two largest equipment loads are the water heater and the dryer:

20 kVA + 5 kVA = 25 kVA

and they draw

Therefore, the 73-A demand load calculated from Table 220.56 satisfies the last sentence of the rule because that value is “not less than” the sum of the largest two kitchen equipment loads. The feeder must be sized to have at least 73 A of capacity for this part of the total building load.

Figure 220-17 shows another example of reduced sizing for a feeder to kitchen appliances.

220.60. Noncoincident Loads. When dissimilar loads (such as space heating and air cooling in a building) are supplied by the same feeder, the smaller of the two loads may be omitted from the total capacity required for the feeder if it is unlikely that the two loads will operate at the same time.

220.61. Feeder Neutral Load. This section covers requirements for sizing the neutral conductor in a feeder, that is, determining the required ampere rating of the neutral conductor. The basic rule of this section says that the minimum required ampacity of a neutral conductor must be at least equal to the “feeder neutral load”—which is the “maximum unbalance” of the feeder load.

“The maximum unbalanced load shall be the maximum net computed load between the neutral and any one ungrounded conductor. . . .” In a 3-wire, 120/240-V, single-phase feeder, the neutral must have a current-carrying

Fig. 220-17. Demand factor for commercial-kitchen feeder. (Sec. 220.56.)

capacity at least equal to the current drawn by the total 120-V load connected between the more heavily loaded hot leg and the neutral. As shown in Fig. 220-18, under unbalanced conditions, with one hot leg fully loaded to 60 A and the other leg open, the neutral would carry 60 A and must have the same rating as the loaded hot leg. Thus No. 6 THW hot legs would require No. 6 THW neutral (copper).

Fig. 220-18. Neutral must be sized the same as hot leg with heavier load. (Sec. 220.61.)

It should be noted that straight 240-V loads, connected between the two hot legs, do not place any load on the neutral. As a result, the neutral conductor of such a feeder must be sized to make up a 2-wire, 120-V circuit with the more heavily loaded hot leg. Actually, the 120-V circuit loads on such a feeder would be considered as balanced on both sides of the neutral. The neutral, then, would be the same size as each of the hot legs if only 120-V loads were supplied by the feeder. If 240-V loads also were supplied, the hot legs would be sized for the total load; but the neutral would be sized for only the total 120-V load connected between one hot leg and the neutral, as shown in Fig. 220-19.

Fig. 220-19. Neutral sizing is not related to phase-to-phase loads. (Sec. 220.61.)

But, there are qualifications on the basic rule of 220.61, as follows:

1. When a feeder supplies household electric ranges, wall-mounted ovens, counter-mounted cooking units, and/or electric dryers, the neutral conductor may be smaller than the hot conductors but must have a carrying capacity at least equal to 70 percent of the current capacity required in the ungrounded conductors to handle the load (i.e., 70 percent of the load on the ungrounded conductors). Table 220.56 gives the demand loads to be used in sizing feeders which supply electric ranges and other cooking appliances. Table 220.55 gives demand factors for sizing the ungrounded circuit conductors for feeders to electric dryers. The 70 percent demand factor may be applied to the minimum required size of a feeder phase (or hot) leg in order to determine the minimum permitted size of neutral, as shown in Fig. 220-20.

Fig. 220-20. Sizing the neutral of a feeder to electric ranges. (Sec. 220.61.)

2. For feeders of three or more conductors—3-wire, DC; 3-wire, single-phase; and 4-wire, 3-phase—a further demand factor of 70 percent may be applied to that portion of the unbalanced load in excess of 200 A. That is, in a feeder supplying only 120-V loads evenly divided between each ungrounded conductor and the neutral, the neutral conductor must be the same size as each ungrounded conductor up to 200-A capacity, but may be reduced from the size of the ungrounded conductors for loads above 200 A by adding to the 200 A only 70 percent of the amount of load current above 200 A in computing the size of the neutral. It should be noted that this 70 percent demand factor is applicable to the unbalanced load in excess of 200 A and not simply to the total load, which in many cases may include 240-V loads on 120/240-V, 3-wire, single-phase feeders or 3-phase loads or phase-to-phase connected loads on 3-phase feeders. Figure 220-21 shows an example of neutral reduction as permitted by 220.61.

WATCH OUT!

The size of a feeder neutral conductor may not be based on less than the current load on the feeder phase legs when the load consists of electric-discharge lighting, data-processing equipment, or similar equipment. The foregoing reduction of the neutral to 200 A plus 70 percent of the current over 200 A does not apply when all or most of the load on the feeder consists of electric-discharge lighting, electronic data-processing equipment, and similar electromagnetic or solid-state equipment. In a feeder supplying ballasts

Fig. 220-21. Neutral may be smaller than hot-leg conductors on feeders over 200 A. (Sec. 220.61.)

for electric-discharge lamps and/or computer equipment, there must not be a reduction of the neutral capacity for that part of the load which consists of discharge light sources, such as fluorescent mercury-vapor or other HID lamps. For feeders supplying only electric-discharge lighting or computers, the neutral conductor must be the same size as the phase conductors no matter how big the total load may be (Fig. 220-22). Full-sizing of the neutral of such feeders is required because, in a balanced circuit supplying ballasts or computer loads, neutral current approximating the phase current is produced by third (and other odd-order) harmonics developed by the ballasts. For large electric-discharge lighting or computer loads, this factor affects sizing of neutrals all the way back to the service. It also affects rating of conductors in conduit because such a feeder circuit consists of four current-carrying wires, which requires application of an 80 percent reduction factor. [See 310.15(B)(2)(a) and 310.15(B)(4)(c).]

In the case of a feeder supplying, say, 200 A of fluorescent lighting and 200 A of incandescent, there can be no reduction of the neutral below the required 400-A capacity of the phase legs, because the 200 A of fluorescent lighting load cannot be used in any way to take advantage of the 70 percent demand factor on that part of the load in excess of 200 A.

It should be noted that the Code wording in 220.61 permits reduction in the size of the neutral when electric-discharge lighting and/or computers are used, if the feeder supplying the electric-discharge lighting load over 200 A happens to be a 120/240-V, 3-wire, single-phase feeder. In such a feeder, the harmonic currents in the hot legs are 180° out of phase with each other and, therefore, would not be additive in the neutral as they are in a 3-phase, 4-wire circuit. In the 3-phase, 4-wire circuit, the third harmonic components of the phase currents are in phase with each other and add together in the neutral instead of canceling out. Figure 220-23 shows a 120/240-V circuit.

Figure 220-24 shows a number of circuit conditions involving the rules on sizing a feeder neutral.

Fig. 220-22. Full-size neutral for feeders to ballast loads or computers. (Sec. 220.61.)

Fig. 220-23. Harmonic loading on true single phase distributions does not interfere with permitted size reductions over 200 A (220.61).

Fig. 220-24. Sizing the feeder neutral for different conditions of loading. (Sec. 220.61.)

Part IV

This part of Art. 220 offers a number of alternative methods for establishing the minimum required current-carrying capacity of service or feeder conductors. Remember that each of the requirements specified within each individual optional method must be satisfied.

220.82. Dwelling Unit. This section sets forth an optional method of calculating service demand load for a residence. This method may be used instead of the standard method as follows:

1. Only for a one-family residence or an apartment in a multifamily dwelling, or other “dwelling unit”

2. Served by a 120/240-V or 120/208-V 3-wire, 100-A or larger service or feeder

3. Where the total load of the dwelling unit is supplied by one set of service-entrance or feeder conductors that have an ampacity of 100 A or greater

This method recognizes the greater diversity attainable in large-capacity installations. It therefore permits a smaller size of service-entrance conductors for such installations than would be permitted by using the load calculations of 220.40 through 220.61.

In making this calculation, as described by 220.82(C), the heating load or the air-conditioning load may be disregarded as a “noncoincident load,” where it is unlikely that two dissimilar loads (such as heating and air conditioning) will be operated simultaneously. In the present NEC, 100 percent of the air-conditioning load is compared with only 40 percent of the total connected load of four or more electric space heaters [220.82(C)(6)], and the lower value is omitted from the calculation. Or, where there are less than four separately controlled electric heating units, the 100 percent value of the air-conditioning load is compared with 65 percent of the load where the electric heating system has less than four separately controlled units.

example A typical application of the data and table of 220.82, in calculating the minimum required size of service conductors, is as follows:

A 1500-sq-ft (139.5-m²) house (excluding unoccupied basement, unfinished attic, and open porches) contains the following specific electric appliances:

12-kW range

2.5-kW water heater

1.2-kW dishwasher

9 kW of electric heat (in five rooms)

5-kW clothes dryer

6-A, 230-V air-conditioning unit

When using the optional method, if a house has air conditioning as well as electric heating, there is recognition in 220.60 that if “it is unlikely that two dissimilar loads will be in use simultaneously,” it is permissible to omit the smaller of the two in calculating required capacity in feeder or in service-entrance conductors. In 220.82, that concept is spelled out in the subparts of 220.82(C) to require adding only the largest of the loads described in this rule. Where the dwelling in question has air conditioning and four separately controlled electric heating units, we add capacity equal to either the total air-conditioning load or 40 percent of the connected load of four or more separately controlled electric space-heating units. For the residence considered here, these loads would be as follows:

Air conditioning = 6 A × 230 V = 1.38 kVA

Note: The air conditioner voltage and current ratings are from the equipment name-plate, and therefore, when converting to kilovoltamperes, must be taken as is. The overall calculation at the end, that determines the service amperage on a 120/240 system, uses the rated system voltages from 220.5(A).

40% of heating (five separate units) = 9 kW × 0.4 = 3.6 kW (3600 VA)

Because 3.6 kW is greater than 1.38 kVA, it is permissible to omit the air-conditioning load and provide a capacity of 3.6 kW in the service or feeder conductors to cover both the heating and air-conditioning loads.

The “other loads” must be totaled up in accordance with 220.82:

In reference to 220.82(C), load categories 1, 2, 3, 4, and 5 are not applicable here: “Air conditioning” has already been excluded as a load because 40 percent of the heating load is greater. The dwelling does not have a heat pump without a controller that prevents simultaneous operation of the compressor and supplemental heating. There is no thermal heating unit. There is no “central” electric space heating; and there are not “less than four” separately controlled electric space-heating units.

The total load of 29,700 VA, as previously summed up, includes “all other load,” as referred to in 220.82. Then:

Using 240- and 120-V values, ampacities may then be calculated. At 240 V, single phase, the ampacity of each service hot leg would then have to be

Minimum service conductor required 100 A

Then the neutral service-entrance conductor is calculated in accordance with 220.61. All 240-V loads have no relation to required neutral capacity. The water heater and electric space-heating units operate at 240 V, 2-wire and have no neutrals. By considering only those loads served by a circuit with a neutral conductor and determining their maximum unbalance, the minimum required size of neutral conductor can be determined.

When a 3-wire, 240/120-V circuit serves a total load that is balanced from each hot leg to neutral—that is, half the total load is connected from one hot leg to neutral and the other half of total load from the other hot leg to neutral—the condition of maximum unbalance occurs when all the load fed by one hot leg is operating and all the load fed by the other hot leg is off. Under that condition, the neutral current and hot-leg current are equal to half the total load watts divided by 120 V (half the volts between hot legs). But that current is exactly the same as the current that results from dividing the total load (connected hot leg to hot leg) by 240 V (which is twice the voltage from hot leg to neutral). Because of this relationship, it is easy to determine neutral-current load by simply calculating hot-leg current load—total load from hot leg to hot leg divided by 240 V.

In the example here, the neutral-current load is determined from the following steps that sum up the components of the neutral load:

Assuming an even balance of this load on the two hot legs, the neutral load under maximum unbalance will be the same as the total load (5100 VA) divided by 240 V (Fig. 220-25) (all results are carried to three significant figures):

Fig. 220-25. Neutral current for lighting and receptacles. The 4.4-A net load equivalence between the laundry circuit and one portion of the lighting circuit is a coincidence. (Sec. 220.82.)

And the neutral unbalanced current for the range load can be taken as equal to the 8000-W range demand load multiplied by the 70 percent demand factor permitted by 220.61 and then divided by 240 V (Fig. 220-26):

Fig. 220-26. Neutral for lighting, receptacles, and range. (Sec. 220.82.)

The clothes dryer contributes neutral load due to the 115-V motor, its controls, and a light. As allowed in 220.61, the neutral load of the dryer may be taken at 70 percent of the load on the ungrounded hot legs. Therefore, the neutral capacity required to accommodate the dryer contribution is (5000 W × 0.7) + 240 V = 14.6 A.

Then, the neutral-current load that is added by the 120-V, 1200-W dishwasher must be added (Fig. 220-27): Although it could be argued that since this load is entirely on a single leg of the system, it should be added to the neutral directly, but that is not the case, as evidenced by the result in Annex D, Example D2(a) from which these numbers are taken. Making the calculation in this way only artificially inflates the size of the neutral. In the real world, the dishwasher and the small-appliance and the laundry and the general lighting circuits all originate from the same panel. There are, in this example and typically, three appliance circuits (laundry and small appliance) all taken at 1500 VA each.

Fig. 220-27. Neutral current for all loads in the example. The lighting load division between line connections is to equalize the load, thereby minimizing the maximum imbalance and the need for a larger feeder neutral. (Sec. 220.82.)

Obviously two will be on one line and one on the other. The number of lighting circuits is unknown but there will be some imbalance there as well. The 1200-W imbalance represented by the dishwasher will be totally lost in the distribution of loads in this panel. Simply placing it against one of the appliance circuits erases its contribution. Further, the 1200-W is not a steady load but one that cycles, depending on whether the booster element is in operation and whether the motor is running. This is why load calculations are usually done on a volt-ampere basis throughout, changing over to amperes only at the point where a conductor size must be determined.

The load calculations for neutral conductors assume reasonable balance for the branch circuits connected to them, as is generally required by 210.11(B). The most meticulously balanced distribution on paper will be defeated by poor panel work. For example, this calculation starts with 21.3 A of line-to-neutral load due to small-appliance and general lighting circuits. If these were arranged to connect to the same line bus, the result would be 42.6 A of current from these sources routinely. And all this could occur before the dishwasher is even connected.

The minimum required neutral capacity is, therefore,

From Code Table 310.16, the neutral minimum for 64 A would be:

4 AWG copper TW or 3 AWG aluminum

6 AWG THW copper* or 4 AWG aluminum*

*If the terminations are evaluated for 75°C connections; and 90°C insulation is permitted provided the conductor size remains as described. Note that if this panel were located remote from the service disconnecting means through a feeder that carried the entire load current, all of these conductors would satisfy the second paragraph of 215.2(A)(1), requiring a reduced neutral to have enough size to carry a line-to-neutral short circuit. The minimum size in this case, per 250.122, is a 8 AWG copper or 6 AWG aluminum.

This book is a handbook based on the 2008 NEC, and the above discussion is the best case the author can make to support the conclusion that is now in the NEC for this example. The method in this example has been unchanged since the 1984 NEC introduced the present 120/240-V nominal voltages, resulting in the recalculation of all the examples. It is also consistent with other neutral calculations throughout the examples, with one notable exception. It is glaringly inconsistent, however, with Example D(1)(b) which has been unchanged for the same amount of time. That example tracks the line contributions of specified 120-V appliances, and the end result shifts accordingly. The approach in Example D(1)(b) can also be supported based on a different reading of the rules in 210.11(B) than the one alluded to above, because the only clear command in that section is to balance loads on circuits that were determined on a load per unit area basis, and a dishwasher load was certainly not part of an area evaluation. Be advised that no clear conclusion can be reached as to how these loads should be calculated at this time.

220.83. Optional Calculation for Additional Loads in Existing Dwelling Unit. This covers an optional calculation for additional loads in an existing dwelling unit that contains a 120/240- or 208/120-V, 3-wire service of any current rating. The method of calculation is similar to that in 220.82.

The purpose of this section is to permit the maximum load to be applied to an existing service without the necessity of increasing the size of the service. The calculations are based on numerous load surveys and tests made by local utilities throughout the country. This optional method would seem to be particularly advantageous when smaller loads such as window air conditioners or bathroom heaters are to be installed in a dwelling with, say, an existing 60-A service, as follows:

If there is an existing electric range, say, 12 kW (and no electric water heater), it would not be possible to add any load of substantial rating. The first 8000 VA is taken at 100 percent, leaving the remainder of permissible load to be calculated at 40 percent. Use the formula 14,400 VA (240 V × 60 A) = 8000 VA + 0.4(X VA), where the quantity (X) is the amount of other load to be evaluated. Rearranging terms gives 6400 VA = 0.4X, so X = 16,000 VA, and therefore, the total “gross load” that can be connected to an existing 120/240-V, 60-A service would be 16,000 VA + 8000 VA = 24,000 VA.

example Thus, an existing 1000-sq-ft dwelling with a 12-kW electric range, two 20-A appliance circuits, a 750-W furnace circuit, and a 60-A service would have a gross load of:

Since the maximum permitted gross load is 24,000 VA, an appliance not exceeding 5250 VA could be added to this existing 60-A service. However, the tabulation at the end of this section lists air-conditioning equipment, central space heating, and less than four separately controlled space-heating units at 100 percent demands; and if the appliance to be added is one of these, then it would be limited even more:

From the 18,750-VA gross load we already have 8000 VA @ 100 percent demand and (10,750 VA [18,750 − 8000] × 0.40) or 4300 VA. The total for the 100 percent and the 40 percent calculation brackets is the sum of 8000 VA and 4300 VA, or 12,300 VA. Then, 14,400 VA (60 × 240 V) − 12,300 VA = 2100 VA for an appliance listed at 100 percent demand.

Although this procedure is limited with respect to saving 60-A services, it can also be applied, with considerably more headroom, to existing 100-A services.

220.84. Multifamily Dwelling. This section provides an optional method of calculating the load in a multifamily dwelling with a fairly high connected load, by reason of electric cooking equipment in all units as well as electric space heating or air conditioning or both. Any house loads are over and above the calculation results from this section, and are to be figured using the standard method in part III. The connected load list for each dwelling unit is formatted the same as the calculation in 220.82, with two differences. There is no 40 percent bracket; instead, all connected loads are simply totaled. In addition, the heating/air-conditioning line is quite simple; just pick the largest number whether the one for heat or the one for air conditioning. Multiply the total per/unit calculation by the number of units, and then by demand factor based on the number of units in Table 220.84.

If the load for a multifamily housing project without electric cooking (and therefore does not initially qualify to use this procedure), as determined by the traditional procedures in part III, turns out to exceed the numbers that come from Table 220.84, the smaller load is permitted to be used.

220.85. Two Dwelling Units. This section provides an optional calculation for sizing a feeder to “two dwelling units.” It notes that if calculation of such a feeder according to the basic long method of calculating given in part III of Art. 220 exceeds the minimum load ampacity permitted by 220.84 for three identical dwelling units, then the lesser of the two loads may be used. This rule was added to eliminate the obvious illogic of requiring a greater feeder ampacity for two dwelling units than for the three units of the same load makeup. Now optional calculations provide for a feeder to one dwelling unit, two dwelling units, or three or more dwelling units.

220.86. Schools. The optional calculation for feeders and service-entrance conductors for a school makes clear that feeders “within the building or structure” must be calculated in accordance with the standard long calculation procedure established by part III of Art. 220. But the ampacity of any individual feeder does not have to be greater than the minimum required ampacity for the whole building, regardless of the calculation result from part III. Note that these calculations differ from most in that they are based on actual load density. The entire connected load is added together, and then divided by the area of the school to generate a load per unit area, whether per square foot or meter. Then the unit load is reduced according to the table, in progressive steps. Finally, the applicable number of volt-amperes per unit area from each step is multiplied by the area of the building to get the final load.

The last sentence in this section excludes portable classroom buildings from the optional calculation method to prevent the possibility that the demand factors of Table 220.86 would result in a feeder or service of lower ampacity than the connected load. Such portable classrooms have air-conditioning loads that are not adequately covered by using a watts-per-square-foot calculation with the small area of such classrooms.

220.87. Determining Existing Loads. Because of the universal practice of adding more loads to feeders and services in all kinds of existing premises, this calculation procedure is given in the Code. To determine how much more load may be added to a feeder or set of service-entrance conductors, at least one year’s accumulation of measured maximum-demand data must be available. Then, the required spare capacity may be calculated as follows:

Additional load capacity = ampacity of feeder or
service conductors − ([1.25 × existing demand
kVA × 1000] ÷ circuit voltage)

where “circuit voltage” is the line-to-line value for single-phase circuits and (1.732) times the phase-to-phase value for 3-phase circuits.

A third required condition is that the feeder or service conductors be protected against overcurrent, in accordance with applicable Code rules on such protection.

If the full-year demand data is not available, an exception allows for a month of monitoring by a continuously recording ammeter, based on maximum demand for the period as defined by average power recorded during 15-min increments. The building must be occupied so the readings will be realistic. In addition, periodic or seasonal loading must be accounted for, either by direct measurements or by calculation, so that the larger of the heating and cooling loads will be included. Although utility demand meter data is commonly available for a full year for service loads, this alternative method is extremely important when the feeder in question is not subject to utility metering, such as a feeder to one part of a building or a feeder connected to a separately derived system.

220.88. New Restaurants. This calculation is available to new restaurants, and produces two different results based on whether or not the restaurant has gas-fired cooking equipment. The numbers for the all-electric restaurants are, of course, significantly higher. The demand table looks somewhat different from comparable tables elsewhere in Art. 220, with one entry providing a 10-percent increment for additional loading over a base number, and others providing far different increments ranging from 20 to 50 percent, and in different relative orders based on the type of restaurant. The table entries are correct.

When this table first came into the NEC, it looked like a conventional table, but it turned out to generate paradoxical results. In one instance adding a few kVA to a load took about 80 A off the ending service calculation, and this was not in just one location. The only way to be sure that additional connected load actually resulted in additional service or feeder capacity was to go back and carefully copyfit demand curves to the utility data that provided the substantiation for the change. That data is well documented, but the resulting curves have some interesting shapes, and the current values in Table 220.88 accurately predict the electrical demand.

Part V. Farm load calculations This part of the article stands alone because farms usually have both a dwelling and a commercial operation connected to a single distribution point. Therefore, some of the loads are eligible for optional treatments in part IV and others are not, requiring a direct transition from the branch circuit calculations in part II to the calculations here. In general, the farmhouse is a dwelling unit and qualifies for the optional dwelling unit procedures in part IV. However, if the dwelling has electric heat and the farm operation uses electric grain drying systems, the dwelling must be calculated under the conventional procedures in part III if it and the barn have a common service.

The electrical equipment for farm operations is taken through Table 220.102, using a master compilation of farm loads in terms of those that will operate continuously and otherwise. Beginning with the continuous loads take the first 60 A at 100 percent and then add the next 60 A of load at 50 percent, and then all other load at 25 percent. Then, for the total farm load, use Table 220.103, which is organized by load. If different buildings or loads have the same function, then those loads can be combined into a single load for these purposes. After the load analysis is complete, add the farm dwelling load calculated as noted above.

ARTICLE 225. OUTSIDE BRANCH CIRCUITS AND FEEDERS

225.1Scope. This article covers all outdoor installations of conductors “on or between buildings, structures, or poles on the premises” and utilization equipment mounted on the outside of buildings, or outdoors on other structures or poles. This rule is followed by the rule of Sec. 225.2, which indicates other Code rules that bear upon the installation of equipment and conductors outdoors on buildings, structures, or poles.

225.3. Part (A) of this section calls for branch circuits to be sized in accordance with the rules of 220.10. And, part (B) of this rule calls for compliance with part III of Art. 220 when sizing outdoor feeder conductors run “on or attached to” buildings, and so forth.

225.4. Conductor Covering. The wiring method known as “open wiring” is recognized in Art. 225 as suitable for overhead use outdoors—“run on or between buildings, structures, or poles” (Fig. 225-1). This is derived from Secs. 225.1, 225.4, 225.14, 225.18, and 225.19. In Sec. 225.4 the Code requires open wiring to be insulated or covered if it comes within 10 ft (3.0 m) of any building or other structure, which it must do if it attaches to the building or structure. Insulated conductors have a dielectric covering that prevents conductive contact with the conductor when it is energized. Covered conductors—such as braided, weatherproof conductors—have a certain mechanical protection for the conductor but are not rated as having insulation, and thus there is no protection against conductive contact with the energized conductor.

Fig. 225-1. Open wiring is OK for overhead circuits. (Sec. 225.4.)

Because 225.4 says that conductors in “cables” (except Type MI) must be of the rubber or thermoplastic type, a number of questions arise.

1. What kind of “cable” does the Code recognize for overhead spans between buildings, structures, and/or poles?

2. May an overhead circuit from one building to another or from lighting fixture to lighting fixture on poles use service-entrance cable, UF cable, or Type NM or NMC nonmetallic-sheathed cable?

The Code covers specific types of cables in turn (Chap. 3), but only in Art. 396 does the Code refer to use for outdoor overhead applications. Effective with the 2008 NEC, 396.30(B) and (C) for the first time cover the use of messenger cable assemblies with the messenger performing an electrical function, thereby closing a long-standing gap in NEC coverage. In addition, the use of service-entrance cable between buildings, structures, and/or poles is supported by Art. 338 in 338.12(A)(3) which points to Part II of Art. 396, thereby including the express reference in 396.10(A).

Use of Type MI, MC, or UF cable for outdoor, overhead circuits is supported by Sec. 396.10. There are no exceptions given to the support requirements in Sec. 334.30 that would let NM or NMC be used aerially, and such cables are not recognized by Sec. 396.10(A) for use as “messenger supported wiring.”

Service-Drop Cable

The NE Code has Art. 396, “Messenger Supported Wiring,” which covers use of “service-drop” cable, but the UL has no listing for or reference to such cable. There is a listing for a suitable medium voltage cable, but traditionally the principal customer for service drop cables has been the electric utilities. And, since they are not usually subject to the NEC because most of their applications are on the line side of the service point, there has not been a large market for a listed product. Article 396 does not require a listing for this product. The NE Code does make reference to it; and its use for aerial circuits between buildings, structures, and/or poles is particularly dictated (Fig. 225-2). Experience with this cable is very extensive and highly satisfactory. It is an engineered product specifically designed and used for outdoor, overhead circuiting.

Fig. 225-2. Aerial cable for overhead circuits. (Sec. 225.4.)

NE Code rules in 230.21 through 230.29 cover use of service-drop cable for overhead service conductors. Because the general rules of Art. 225 on outside branch circuits and feeders do make frequent references to other sections of Art. 230, it is logical to equate cables for overhead branch circuits and feeders to cables for overhead services. Although the rules of Art. 396 refer to a variety of messenger-supported cable assemblies, for outdoor circuits, use of service-drop cable is the best choice—because such cable is covered by the application rules of 230.21 through 230.29. Other types of available aerial cable assemblies, although not listed by UL, might satisfy some inspection agencies. But, in these times of OSHA emphasis on codes and standards, use of service-drop cable has the strongest sanction.

One important consideration in the use of service-drop cable as a branch circuit or feeder is the general Code prohibition against use of bare circuit conductors. 310.2 requires conductors to be insulated. An exception notes that bare conductors may be used where “specifically permitted.” Bare equipment grounding conductors are permitted in 250.118. A bare conductor for SE cable is permitted in 338.100. Bare neutrals are permitted for service-entrance conductors in the Exception to 230.41, for underground service-entrance (service lateral) conductors in the Exception to 230.30, and for service-drop conductors in 230.22, Exception when used as service conductors. When service-drop cable is used as a feeder or branch circuit, however, there is no permission for use of a bare circuit conductor—although it may be acceptable to use the bare conductor of the service-drop cable as an equipment grounding conductor. And where service-drop cable is used as a feeder from one building to another, it would seem that a bare neutral could be acceptable as a grounded neutral conductor—as permitted in the last sentence of 338.3(B), first paragraph. This is where the new language in 396.30 is useful, because it expressly recognizes this use when it complies with 225.4, and 225.4 Exception allows the bare neutral where recognized elsewhere, such as where a regrounded neutral is permitted as covered in 250.32(B) Exception. Note that such regrounded neutrals are only permitted for existing premises wiring systems, so this use will gradually disappear.

When service-drop cable is used between buildings, the method for leaving one building and entering another must satisfy 230.52 and 230.54. This is required in Sec. 225.11.

The Exception to 225.4 excludes equipment grounding conductors and grounded circuit conductors from the rules on conductor covering. This Exception permits equipment grounding conductor and grounded circuit conductors (neutrals) to be bare or simply covered (but not insulated) as permitted by other Code rules.

Because the matter of outdoor, overhead circuiting is complex, check with local inspection agencies on required methods. As NE Code 90.4 says, the local inspector has the responsibility for making interpretations of the rules.

225.5. Size of Conductors 600-V Nominal or Less. This rule calls for conductor ampacity to be determined in accordance with 310.15 and the rules of Art. 220. But remember, where the load to be supplied is continuous or a combination of continuous and noncontinuous loads, then the rules of 210.19(A) and 210.20(A) or 215.2(A) and 215.3(A), covering conductor sizing and OC protection for branch circuits and feeders supplying continuous loads, must be observed, as well.

225.6. Minimum Size of Conductor. Open wiring must be of the minimum sizes indicated in 225.6 for the various lengths of spans indicated.

Article 100 gives a definition of festoon lighting as “a string of outdoor lights suspended between two points” (Fig. 225-3). Such lighting is used at carnivals, displays, used-car lots, etc. Such application of lighting is limited because it has a generally poor appearance and does not enhance commercial activities.

As covered in 225.6(B), overhead conductors for festoon lighting must not be smaller than No. 12; and where any span is over 40 ft (12.0 m), the conductors must be supported by a messenger wire, which itself must be properly secured to strain insulators. But the rules on festoon lighting do not apply to overhead circuits between buildings, structures, and/or poles.

Fig. 225-3. Festoon lighting is permitted outdoors. (Sec. 225.6.) Note that 590.4(F) requires a guard on lampholder such as the one shown here if the use qualifies as temporary wiring.

225.7. Lighting Equipment Installed Outdoors. Part (B) permits a common neutral for both outdoor branch circuits and feeders—something not permitted for indoor branch circuits (a neutral of a 3-phase, 4-wire circuit is not a common neutral), although 215.4 grants limited permission for feeders with common neutrals. For two 208Y/120-V multiwire circuits consisting of six ungrounded conductors (two from each phase) and a single neutral (serving both circuits) feeding a bank of floodlights on a pole, if the maximum calculated load on any one circuit is 12 A and the maximum calculated load on any one phase is 24 A, the ungrounded circuit conductors may be No. 14, but the neutral must be at least No. 10. This rule clearly states the need to size a common neutral for the maximum (most heavily loaded) phase leg made up by multiple conductors connected to any one phase and supplying loads connected phase-to-neutral.

Part (C) covers use of 480/277-V systems for supplying incandescent and electric-discharge lighting fixtures. This section rules that outdoor fixtures installed for lighting “outdoor areas” at commercial or public buildings must be not less than 3 ft (900 mm) from “windows, platforms, fire escapes, and the like.”

225.10. Wiring on Buildings. This section identifies those wiring methods, rated up to 600 V, that are permitted to be mounted on the exterior of buildings. Note that rigid nonmetallic conduit may be used for outside wiring on buildings, as well as the other raceway and cable methods covered in this section. For a long time, rigid PVC was not permitted for such application. Installation of conductors rated over 600 V must comply with the provisions of 300.37, and electric signs and outline lighting must be installed as dictated by the rules of Art. 600.

225.14. Open-Conductor Spacings. Open wiring runs must have a minimum spacing between individual conductors (as noted in 225.14) in accordance with Table 230.51(C), which gives the spacing of the insulator supports on a building surface and the clearance between individual conductors on the building or where run in spans (Fig. 225-4).

Fig. 225-4. Spacing of open-wiring conductors. (Sec. 225.14.)

It should be noted that 225.14 and Table 230.51(C) require that the minimum spacing between individual conductors in spans run overhead be 3 in. (76 mm) for circuits up to 300 V (such as 120, 120/240, 120/208, and 240 V). For circuits up to 600 V, such as 480 and 480/277 V, the minimum spacing between individual conductors must be at least 6 in. (152 mm).

225.17. Masts as Supports. Masts must have sufficient rigidity to handle the strain, or they must be guyed accordingly. If a raceway mast is used, any fittings must be identified for use with masts, and only the feeder and branch-circuit conductors within the scope of Art. 225 can be attached to the mast. For example, a telephone drop is not permitted to be attached to a power circuit mast, regardless of the strength of the raceway or the amount of guy wire support provided.

225.18. Clearance from Ground. Overhead spans of open conductors and open multiconductor cables must be protected from contact by persons by keeping them high enough above ground or above other positions where people might be standing. And they must not present an obstruction to vehicle passage or other activities below the lines (Fig. 225-5).

The rule of 225.18 applies to “open conductors and open multiconductor cables” and gives the conditions under which clearances must be 10, 12, 15, or 18 ft (3.0, 3.7, 4.5, or 5.5 m)—for conductors that make up either a branch circuit or a feeder [not service-drop conductors, which are subject to 230.24(B)]. Although the wording used here is the same as that referring to corresponding clearances in 230.24(B), 225.18 covers those “open conductors and open multi-conductor cables”—such as triplex and quadruplex cables—that do not meet the definition of service conductors, which would be regulated by Art. 230. Article 225 gives minimum clearances for triplex or quadruplex cables, as well as open individual conductors, commonly used for outdoor overhead branch circuits and feeders.

As 225.18 stands, “open conductors and multiconductor cables” for an overhead branch circuit or feeder require only a 10-ft (3.05-m) clearance from ground for circuits up to 150 V to ground; just as service-drop conductors up to 150 V must have a clearance of not less than 10 ft (3.05 m) from ground.

Fig. 225-5. Conductor clearance from ground. (Sec. 225.18.)

The rules of this section agree with the clearances and conditions set forth in the NESC (National Electrical Safety Code) for open conductors outdoors. The distances given for clearance from ground must conform to maximum voltage at which certain heights are permitted.

225.19. Clearances from Buildings for Conductors of Not Over 600 V, Nominal. The basic minimum required clearance for outdoor conductors running above a roof is 8-ft (2.5-m) vertical clearance from the roof surface.

The basic ideas behind the rules are as follows:

1. Any branch-circuit or feeder conductors—whether insulated, simply covered, or bare—must have a clearance of at least 8 ft (2.5 m) vertically from a roof surface over which they pass. And that clearance must be maintained not less than 3 ft (900 mm) from the edge of the roof in all directions.

2. A roof that is subject to “pedestrian or vehicular traffic” must have conductor clearances “in accordance with the clearance requirements of 225.18.” That reference essentially requires a clearance of 12 ft above a roof that serves as driveway or parking area, not subject to “truck traffic,” and where the voltage to ground does not exceed 300 V to ground. Where the voltage to ground exceeds 300 V to ground, then a minimum clearance of 15 ft must be provided. And, if the area is subject to truck traffic and the conductors are operated at more than 300 V to ground, then a minimum of clearance above the roof of 18 ft must be provided.

In parts (B) and (C), overhead conductor clearance from signs, chimneys, antennas, and other nonbuilding or nonbridge structures must be at least 3 ft (900 mm)—vertically, horizontally, or diagonally.

Part (D) addresses installation details regarding clearance of “final spans.” In part (D)(1), the Code requires that the connection point of overhead branch circuit and feeder conductors to the building be kept at least 3 ft from any of the building openings identified by the first part of this rule. The rule exempts windows that do not open from compliance.

But in part (D)(2) the Code addresses those final spans that run above areas that people may occupy. Where the final span’s connection point runs above, or is within 3 ft horizontally from, “platforms, projections, or surfaces” where a person could come into contact with the conductors or cable, the clearances given in 225.18 must be observed.

Part (D)(3) prohibits installation of outside branch-circuit and feeder conductors beneath, or where they obstruct the entrance to, building openings through which material or equipment is intended to be moved. Barns provide a good example of the type of building opening this rule is intended to cover. Although only “farm and commercial buildings” are mentioned, they are only held up as examples. The wording used extends this requirement to any such opening, at any occupancy.

As indicated in Fig. 225-6, Exception No. 2 to 225.19(A) may apply to circuits that are operated at 300 V or less.

Fig. 225-6. Conductors—whether or not they are fully insulated for the circuit voltage—must have at least 8-ft (2.44-m) vertical clearance above a roof over which they pass. (Sec. 225.19.)

Part (E) covers a preferred exclusion zone (required only if practicable) in which overhead lines should not be run adjacent to high-rise buildings in order that fire ladders can be set up.

225.22. Raceways on the Exterior Surfaces of Buildings or Other Structures. Condensation of moisture is very likely to take place in conduit or tubing located outdoors. The conduit or tubing should be considered suitably drained when it is installed so that any moisture condensing inside the raceway or entering from the outside cannot accumulate in the raceway or fittings. This requires that the raceway shall be installed without “pockets,” that long runs shall not be truly horizontal but shall always be pitched, and that fittings at low points be provided with drainage openings.

In order to be raintight, all conduit fittings must be made up wrench-tight. Couplings and connectors used with electrical metallic tubing shall be listed as “raintight.”

225.24. Outdoor Lampholders. This section applies particularly to lampholders used in festoons. Where “pigtail” lampholders are used, the splices should be staggered (made a distance apart) in order to avoid the possibility of short circuits, in case the taping for any reason should become ineffective.

According to the UL Standard for Edison-Base Lampholders, “pin-type” terminals shall be employed only in lampholders for temporary lighting or decorations, signs, or specifically approved applications. The NEC requires that such lampholders only be used on stranded wire.

225.25. Location of Outdoor Lamps. In some types of outdoor lighting it would be difficult to keep all electrical equipment above the lamps, and hence a disconnecting means may be required. A disconnecting means should be provided for the equipment on each individual pole, tower, or other structure if the conditions are such that lamp replacements may be necessary while the lighting system is in use. It may be assumed that grounded metal conduit or tubing extending below the lamps would not constitute a condition requiring that a disconnecting means must be provided.

225.26. Vegetation As Support. Trees or any other “vegetation” must not be used “for support of overhead conductor spans.” Note that the wording used here does not include electric equipment, but, rather, only prohibits “overhead conductor spans” from being supported by “vegetation.” The effect is to permit outdoor lighting fixtures to be mounted on trees and to be supplied by an approved wiring method—conductors in a raceway or Type UF cable—attached to the surface of the tree (Fig. 225-7).

Fig. 225-7. The rule of 225.26 prohibits wiring installed on trees.

Part II. More Than One Building or Structure

225.30. Number of Supplies. These rules cover those installations where several buildings are supplied from a single service. Although technically “feeder” or “branch-circuit” conductors, given that the supply conductors to the other buildings are the effective equivalent of service conductors, the rules given in part II of Art. 225 are very nearly identical to those given for service conductors by Art. 230. In fact, many of the rules here were simply lifted from Art. 230 and modified as needed for use in Art. 225.

In 225.30, the Code stipulates the number of sources of supply to one building from another. The basic rule is similar to 230.2 for services, which calls for no more than one source of supply. Of course, as with 230.2, the rules in 225.30(A) through (E) present a number of circumstances where it would be permissible to supply one building or structure from another with more than one source. The wording used here is identical to that used in Sec. 230.2, which regulates the number of services permitted, simply because the “feeder” from the main building (i.e., the one where the service is installed) is essentially or effectively the “service” to the second building. For a group of buildings under single management, disconnect means must be provided for each building, as in Fig. 225-8.

Fig. 225-8. Each building must have its own disconnect means. (Sec. 225.32.)

The rules for services will be exhaustively covered in Art. 230, so this discussion will only focus on the differences between parts II and III of this article and the service article. The principal general difference is that wiring from one building to another, although superficially like service wiring, is not service wiring. A service entrance is the interface between premises wiring and the facilities of the serving utility. It is also the interface between wiring governed by the National Electrical Safety Code (NESC) and premises wiring governed by the National Electrical Code. These codes are very different because the NESC presumes that the maintenance of those systems will be performed by a highly regulated utility workforce operating under a unique workplace culture and environment, and that this will continue for the foreseeable future, given the organizational permanence of utility enterprises. If there is no utility interface, then there is no service. For this reason, it is incorrect to label the disconnecting means for one building fed from another as a “service” disconnect when it is actually a “building” disconnect. The real significance of the creation of this part of Art. 225 is the clarity it brings to what is encompassed by a service, and more importantly, what is not. Specific differences are as follows:

As previously noted, only one source of supply is permitted as a general rule. This is the same general rule as for services, but there are subtle differences. The permission to run multiple service laterals from a common connection to disconnected loads does not apply. The allowance for multiple supplies where the capacity requirement exceeds that which the local utility supplies through one service does not apply for obvious reasons, and the allowance for an exemption on capacity grounds by special permission does not apply either. The final difference is the permission, granted here and not for services, for additional supplies where there are “documented switching procedures” in place for safe disconnection. This normally arises on large, campus-style industrial distributions.

225.31. Disconnecting Means. The first sentence of this rule mandates the installation of a disconnecting means to permit the feeder or branch-circuit conductors that supply “or run through” a building to be deenergized.

The location and other details are presented by 225.32.

As could be anticipated, the required disconnecting means for the building supply conductors must be installed at a “readily accessible location” as defined in Art. 100 of Chap. 1.

225.32 calls for the disconnecting means to be located at the point where the supply conductors enter the building if the disconnecting means are located inside or outside the building or structure served. This rule is hard to understand. The wording here is essentially the same as it is for service conductors. But, while service conductors are unprotected, these feeder or branch-circuit conductors must have OC protection at their supply end! Why mandate the disconnecting means to be installed immediately at the point of conductor entry? There doesn’t seem to be any good answer, but nonetheless, observation of this rule is mandatory. Where installed inside an auxiliary building or structure, supplied from a service in another building, the disconnecting means required by this rule must be at the point of conductor entry.

Note that this rule applies to any building fed from another, even a detached garage for a single-family house. The clear intent is that it should never be necessary to enter a second building to disconnect the one at hand, with rare and very specific exceptions.

The last sentence of 225.32 states that the remedies provided in 230.6 apply to the feeder and branch-circuit conductors supplying an outbuilding. This allowance can be used, just as for service conductors, to artificially extend the point of entry into a building to an interior disconnecting means.

Exception No. 1 applies to commercial, industrial, and institutional occupancies where a full-time staff provides maintenance. If such a facility’s maintenance staff have established—in writing—“safe switching procedures,” then the disconnecting means required by the basic rule may be located remotely from the building supplied.

For any “integrated electrical system” as defined and regulated by Art. 685, Exception No. 2 suspends the basic rule calling for a feeder disconnect at each building.

Exception No. 3 eliminates the need for individual disconnects for individual lighting standards. The literal wording calls for a disconnect at each “structure.” The addition of this Exception indicates the CMP’s intent, which is to permit one disconnect for a number of lighting poles. And Exception No. 4 extends similar recognition to “poles or similar structures” that support signs (Art. 600).

225.33. Maximum Number of Disconnects. The six-disconnect rule is almost the same as for services, but in this application, disconnects for surge-protective equipment and power monitoring equipment are not exempted from the allowable total of six. The access requirements, the grouping requirement, the requirement to segregate disconnects for certain critical systems, and the specification of minimum ratings for certain applications are the same as for services.

225.36. Suitable as Service Equipment. This section requires that the disconnect for each building of a multibuilding layout be recognized for service use—usually that means that the disconnecting means for each building must be listed by UL, or another national test lab, as suitable for service equipment. This means the building disconnect is supplied as described in 250.32(B) Exception, which covers grounding and bonding requirements for an outbuilding main disconnect where the neutral, without an equipment grounding path, is run to the outbuilding disconnect. Where an outbuilding main disconnect is so supplied and grounded or bonded, it must be suitable for use as service equipment. There is no exception covering the more usual case where the out-building main disconnect is supplied with both an equipment ground and a neutral (grounded) conductor. Therefore in such cases the equipment rating requirement still stands, but the neutral-to-ground bonding means provided with the disconnect must not be used. The other rules regarding how the disconnecting means are to be constructed (including the simultaneous opening of poles, the special provisions for a grounded circuit conductor, the method of operation, and the indication requirement) are all the same as for services.

The Exception recognizes the use of wiring device switches (snap switches) as the required disconnecting means at garages and outbuildings on residential property. However, each disconnecting means still must be grouped and marked to indicate its function and the load served. A further residential compromise involves a waiver of the reciprocal labeling rule when multiple circuits supply a dwelling. The Code-making panel decided that reciprocal signage in such a building to the effect of “This is disconnect 1 of 2, controlling the overhead light; disconnect 2 of 2 located on the west side of the garage door controls the GFCI receptacle” would be excessive. The reciprocal labeling waiver does not, however, waive the identification rule on each switch; nor does it waive any other disconnecting requirements covered here. Specifically:

1. Ungrounded conductors supplying a load intended to stay energized, such as a receptacle, must pass through a disconnecting means located at a readily accessible point nearest the point of entrance.

2. A snap switch is a permissible disconnecting means, including a three-way switch with no identifiable off position.

3. The switches associated with a single source of supply, such as a single branch circuit, must be grouped, although they needn’t be as close as adjacent snap switches in a two-gang box.

4. Each switch must be marked with its function. If that function is obvious, such as the overhead light, NEC 110.22 allows some basis for omitting this marking. However, by providing the marking you will avoid challenges.

Suppose you install a receptacle that will supply a freezer, and the owner wants to be assured that it won’t be turned off inadvertently. Assume there will also be a light controlled from the house and the garage using three-way switches. Mark the three-way switch in the garage LIGHT. Run the receptacle feed through a single-pole snap switch in another box near the three-way switch, perhaps at an odd height, say 3 ft above the floor. Over the other switch place a weatherproof cover that precludes inadvertent operation, and mark it RECEP DISC. or similar.

Part III, Over 600 V This part covers medium voltage supply wiring (over 600 V). Most of this part, as in the case of part II, matches up with the equivalent requirements for services, and only the differences will be covered here.

1. Since circuits originating in other buildings will necessarily be either feeders or branch circuits, and not service conductors, they will have overcurrent protection at their supply ends and need simply be sized for the load according to the procedures for comparable circuits set out in the NEC articles covering branch circuits and feeders. Medium voltage service disconnects are often pole-top devices at the edge of a property, and less amenable to a readily access than a building disconnect. The requirements for isolating switches, the construction details of the disconnecting means, and the allowances for remote control operation are the same as for services.

2. Medium-voltage feeders and branch circuits run overhead need to meet enhanced clearances reflecting the increased hazard involved for circuits operating over 600 V. These conductors must be installed so that they are at least 2.3 m (7.5 ft) away (horizontally separated) from building walls, projections, and windows. They must observe the same horizontal spacing from balconies, catwalks, and similar areas that people would have access to, and the same distance applies to other structures.

3. The same conductors must be at least 3.8 m (12.5 ft) either above a roof or below a roof (or other projection) where run at that level. For roofs accessible to vehicles (but not truck traffic) such as parking garages, the vertical clearance rises to 4.1 m (13.5 ft), 5.6 m (18.5 ft) if truck traffic uses the roof. Clearances over open ground vary by the type of traffic as well, beginning at 4.1 m (12.5 ft) for walkways and then rising a foot to 4.4 m (14.5 ft) for pedestrian ways and restricted traffic. Water areas not suitable for boating come in at 5.2 m (17 ft) and then open land suitable for grazing, cultivation, or vehicles, along with ways subject to vehicular access generally, including roads, driveways, alleys, and parking lots, all at 5.6 m (18.5 ft). The highest prescribed clearance, 8.1 m (26.5 ft), applies to runs over railways. These clearances all apply to medium voltages up to and including 22 kV measured to ground. Higher voltages add 10 mm (0.4 in) per kilovolt above 22 kV, and special cases, including clearances over navigable waters, and areas with large vehicles such as mining operations may require special engineering and review by the authority having jurisdiction.

ARTICLE 230. SERVICES

230.2. Number of Services. For any building, the service consists of the conductors and equipment used to deliver electric energy from the utility supply lines to the interior distribution system. Service may be made to a building either overhead or underground, from a utility pole line or from an underground transformer vault.

The first sentence of this rule requires that a building or structure be supplied by “only one service.” Because “service” is defined in Art. 100 as “The conductors and equipment for delivering energy from the serving utility to the wiring system of the premises served,” use of one “service” corresponds to use of one “service drop” or one “service lateral.” Thus, the basic rule of this section requires that a building or other structure be fed by only one service drop (overhead service) or by only one set of service lateral conductors (underground service). As shown in Fig. 230-1, a building with only one service drop to it satisfies the basic rule even when more than one set of service-entrance conductors are tapped from the single drop (or from a single lateral circuit). Also note that only a utility may supply a service. A power source consisting of a generator, or even an on-site electric plant, is a separately derived system and the applicable rules fir disconnects, etc., will be found in Part II of Art. 225 and not this article. And when such energized conductors reach the premises in question, they will pass through a “building disconnect” and not a “service disconnect.” Review the coverage at the beginning of Part II of Art. 225 for more information on this crucial topic.

230.2adds an important qualification of that rule as it applies only to 230.40, Exception No. 2, covering service-entrance layouts where two to six service disconnects are to be fed from one drop or lateral and are installed in separate individual enclosures at one location, with each disconnect supplying a separate load. As described in 230.40, Exception No. 2, such a service equipment layout may have a separate set of service-entrance conductors run to “each or several” of the two to six enclosures. The second sentence in 230.2 notes that

Fig. 230-1. One set of service-drop conductors supply building from utility line (coming from upper left) and two sets of SE conductors are tapped through separate metering CTs. (Sec. 230.2.)

where a separate set of underground conductors of size 1/0 or larger is run to each or several of the two to six service disconnects, the several sets of underground conductors are considered to be one service (i.e., one service lateral) even though they are run as separate circuits, that is, connected together at their supply end (at the transformer on the pole or in the pad-mount enclosure or vault) but not connected together at their load ends. The several sets of conductors are taken to be “one service” in the meaning of 230.2, although they actually function as separate circuits (Fig. 230-2).

Although 230.40, Exception No. 2, applies to “service-entrance conductors” and service equipment layouts fed by either a “service drop” (overhead service) or a “service lateral” (underground service), the second sentence in 230.2 is addressed specifically and only to service “lateral” conductors (as indicated by the word “underground”) because of the need for clarification based on the Code definitions of “service drop,” “service lateral,” “service-entrance conductors, overhead system,” and “service-entrance conductors, underground system.” (Refer to these definitions in the Code book to clearly understand the intent of this part of 230.2 and its relation to 230.40, Exception No. 2.)

The matter involves these separate but related considerations:

1. Because a “service lateral” may (and usually does) run directly from a transformer on a pole or in a pad-mount enclosure to gutter taps where short tap conductors feed the terminals of the service disconnects, most layouts of that type literally do not have any “service-entrance conductors” that would be subject to the application permitted by 230.40, Exception No. 2—other than the short lengths of tap conductors in the gutter or box where splices are made to the lateral conductors.

Fig. 230-2. “One” service lateral may be made up of several circuits. (Sec. 230.2.)

2. Because 230.40, Exception No. 2, refers only to sets of “service-entrance conductors” as being acceptable for individual supply circuits tapped from one drop or lateral to feed the separate service disconnects, that rule clearly does not apply to “service lateral” conductors which by definition are not “service-entrance conductors.” So there is no permission in 230.40, Exception No 2, to split up “service lateral” capacity. And the first sentence of 230.2 has the clear, direct requirement that a building or structure be supplied through only one lateral for any underground service. That is, either a service lateral must be a single circuit of one set of conductors, or if circuit capacity requires multiple conductors per phase leg, the lateral must be made up of sets of conductors in parallel—connected together at both the supply and load ends—in order to constitute a single circuit (i.e., one lateral).

3. 230.2 permits “laterals” to be subdivided into separate, nonparallel sets of conductors in the way that 230.40, Exception No. 2, permits such use for “service-entrance conductors”—but only for conductors of 1/0 and larger and only where each separate set of lateral conductors (each separate lateral circuit) supplies one or several of the two to six service disconnects.

230.2recognizes the importance of subdividing the total service capacity among a number of sets of smaller conductors rather than a single parallel circuit (i.e., a number of sets of conductors connected together at both their supply and load ends). The single parallel circuit would have much lower impedance and would, therefore, require a higher short-circuit interrupting rating in the service equipment. The higher impedance of each separate set of lateral conductors (not connected together at their load ends) would limit short-circuit current and reduce short-circuit duty at the service equipment, permitting lower AIR (ampere interrupting rating)-rated equipment and reducing the destructive capability of any faults at the service equipment.

Subparts (A) through (E) cover cases where two or more service drops or laterals may supply a single building or structure.

230.2(A) permits a separate drop or lateral for supply to a fire pump and/or to emergency electrical systems, such as emergency lighting or exit lights and/or standby systems.

Part (A), which is essentially an exception to the basic rule that a building “shall be supplied by only one service,” also recognizes use of an additional power supply to a building from any “parallel power production systems.” This would permit a building to be fed by a solar photovoltaic, wind, or other electric power source—in addition to a utility service—just as an emergency or standby power source is also permitted (Fig. 230-3). Fire pumps may be supplied from a separate service drop or lateral, and service from optional standby

Fig. 230-3. Electric power generated by a solar voltaic assembly or by a wind-driven generator may be used as a source of power in “parallel” with the normal service. (Sec. 230.2.)

systems, legally required standby systems, and emergency systems are permitted in addition to the “one” service required by the basic rule in 230.2, “Number of Services.” The final condition covered in this part is the “system designed for connection to multiple sources of supply for the purpose of enhanced reliability.” This is the widely used double-ended switchboard with services provided to each end, and some form of throw-over in the middle so if one end goes down, the other end can supply the entire occupancy.

In 230.2(B) the Code recognizes other situations in which more than one service (i.e., more than one service drop or lateral) may be used. By “special permission” of the inspection authority, more than one service may be used for a multitenant building when there is no single space that would make service equipment available to all tenants.

Part (B)(2) requires special permission to install more than one service to buildings of large area. Examples of large-area buildings are high-rise buildings, shopping centers, and major industrial plants. In granting special permission the authority having jurisdiction must examine the availability of utility supplies for a given building, load concentrations within the building, and the ability of the utility to supply more than one service. Any of the special-permission clauses in 230.2 require close cooperation and consultation between the authority having jurisdiction and the serving utility. And, as always, such “special permission” must be provided in writing to satisfy the wording of the definition for “special permission” given in Art. 100.

Two or more services to one building are also permitted by part (C) when the total demand load of all the feeders is more than 2000 A, up to 600 V, where a single-phase service needs more than one drop, or by special permission (Fig. 230-4). 230.2(C) relates capacity to permitted services. Where requirements exceed 2000 A, two or more sets of service conductors may be installed. Below this value, special permission is required to install more than one set. The term “capacity requirements” appears to apply to the total calculated load for sizing service-entrance conductors and service equipment for a given installation, which would mean that the load calculated in accordance with Art. 220 must exceed 2000 A before one can assume permission for more than one set of service conductors.

Cases of separate light and power services to a single building and separate services to water heaters for purposes of different rate schedules are also permitted. And if a single building is so large that one service cannot handle the load, special permission can be given for additional services.

230.2(D) is illustrated at the bottom of Fig. 230-4.

The last part of the rules in 230.2, part (E), introduces a requirement that applies to any installation where more than one service is permitted by the Code to supply one building. It requires a “permanent plaque or directory” to be mounted or placed “at each service drop or lateral or at each service-equipment location” to advise personnel that there are other services to the premises and to tell where such other services are and what building parts they supply.

This directory must be placed at each service. So, if there are two services, there should be two plaques at each service location. The directory (or directories) must identify all feeders and branch circuits supplied from that service.

Fig. 230-4. Exceptions to 230.2 permit two or more services under certain conditions. (Sec. 230.2.)

Further, such directories must be fully reciprocal both as to load descriptions and locations. That is, if plaque number one says: “This is Service #1 of 2, for the north half of the building. Service #2, located in the middle of the south wall, is for the south end of the building” then there must be another such plaque in the middle of the south end of the building, reading something like this: “This is Service #2 of 2, for the south half of the building. Service #1, located in the middle of the north end of the building, is for the north end of the building.” Labeling that provides this type of fully reciprocal information is required at every service equipment location if more than one service arrives at the building, for whatever reason. In addition, in some cases a building will be fed directly by a service and also by a feeder from another building or perhaps from a separately derived system. The reciprocal labeling rules apply to those supply systems just as if they were services.

230.3. One Building or Other Structure Not to Be Supplied Through Another. For the most part, the service conductors supplying each building or structure shall not pass through the inside of another building. The concern here is related to the fact that service-entrance conductors have no overcurrent protection at their line end. They are simply connected to the utility’s supply without any type of OC device. Although the utility may have fuses in its lines, the fuses probably won’t open the circuit unless there is a bolted-fault, which represents the smallest percentage of all faults. Effectively speaking, this means that any other fault in the service conductor is expected to “burn” itself clear. That being the case, the Code here prohibits running unprotected service conductors through one building to another, unless they are in a raceway encased by 2 in. (50 mm) of concrete or masonry (Fig. 230-5). 230.6 points out that conductors in a raceway enclosed within 2 in. (50 mm) of concrete or masonry are considered to be “outside” the building even when they are run within the building.

Fig. 230-5. This is not a violation of the basic rule of 230.3. (Sec. 230.3.)

A building as defined in Art. 100 is a “structure which stands alone or which is cut off from adjoining structures by fire walls with all openings therein protected by approved fire doors.” A building divided into four units by such fire walls may be supplied by four separate service drops, but a similar building without the fire walls may be supplied by only one service drop, except as permitted in 230.2.

A commercial building may be a single building but may be occupied by two or more tenants whose quarters are separate, in which case it might be undesirable to supply the building through one service drop. Under these conditions special permission may be given to install more than one service drop.

230.6. Conductors Considered Outside of Building. A complement to the requirement in 230.3, this section presents certain criteria that, when satisfied, render the service equipment and/or conductors “outside” the building. For example, in part (1), the Code states that service conductors are considered “outside” the building when run in conduit under a building, and the conduit is covered with either brick or concrete at least 2 in. thick.

In part (2) the Code recognizes that conductors in conduit or duct enclosed by concrete or brick not less than 2 in. (50 mm) thick are considered to be outside the building, even though they are actually run within the building. Figure 230.6 shows how a service conduit was encased within a building so that the conductors are considered as entering the building right at the service protection and disconnect where the conductors emerge from the concrete, to satisfy the rule of 230.70(A), which requires the service disconnect to be as close as possible to the point where the SE conductors enter the building. Figure 230-7 shows an actual case of this application, where forms were hung around the service conduit and then filled with concrete to form the required concrete case.

Part (3) considers service equipment and conductors, installed within a fire-rated vault that conforms with the Code rules for transformer vaults given in part III of Art. 450, to be outside the building. Part (4) presents recognition similar to that given in part (1), but this covers conductors buried in raceways under at least 18 in. of earth.

Fig. 230-6. “Service raceways” in concrete are considered “outside” a building. (Sec. 230.6.)

230.7. Other Conductors in Raceway or Cable. Although the basic rule permits only service-entrance conductors to be used in a service raceway or service cable, exceptions do recognize the use of grounding conductors in a service raceway or cable and also permit conductors for a time switch if overcurrent

Fig. 230-7. Top photo shows service conduit carried above suspended ceiling, without the SE disconnect located at the point of entry. When conduit was concrete-encased, the service conductors then “enter” the building at the SE disconnect—where they emerge from the concrete. Service conduit enters building at lower left and turns up into SE disconnect (right) in roof electrical room.

protection is provided for the conductors, as shown in Fig. 230-8. Refer to the discussion of the definition of “raceway” in Art. 100 (Chap. 1 in this book) for more information regarding the status of auxiliary gutters with respect to this rule and why, for this reason, they are not classified as raceways.

Fig. 230-8. A time switch with its control circuit connected on the supply side of the service equipment. (Sec. 230.7.)

230.8. Raceway Seal. Figure 230-9 indicates that Sec. 300.5(G) applies to underground service conduits. Where service raceways are required to be sealed—as where they enter a building from underground—the sealing compound used must be marked on its container or elsewhere as suitable for safe and effective use with the particular cable insulation, with the type of shielding used, and with any other components it contacts. Some sealants attack certain insulations, semiconducting shielding layers, and so forth.

Fig. 230-9. Service raceways may have to be sealed. (Sec. 230.8.)

230.9. Clearance from Building Openings. Parts (A), (B), and (C) cover the clearance requirements for service conductors, including the final portion of overhead spans and their point of connection to the building or structure.

In Part (A), the Code makes clear that any service-drop conductors—open wiring or multiplex drop cable—must have the 3-ft (900-mm) clearance from windows, doors, porches, and so forth, to prevent mechanical damage to and accidental contact with service conductors (Figs. 230-10 and 230-11). The clearances required in 230.24, 230.26, and 230.29 are based on safety-to-life

Fig. 230-10. Drop conductors must have clearance from building openings. (Sec. 230.9.)

Fig. 230-11. Drop conductors above top level of a window or door do not require 3-ft horizontal clearance. (Sec. 230.9, Exception.)

considerations in that wires are required to be kept a reasonable distance from people who stand, reach, walk, or drive under service-drop conductors. As the Exception notes, conductors that run above the top level of a window do not have to be 3 ft (900 mm) away from the window.

The rule of 230.9(B) recognizes clearances of less than 3 ft horizontally from porches, balconies, and so on, provided the minimum vertical clearance, measured from the floor of the porch or balcony, is in accordance with 230.24(B). The service-drop conductors shown in the drawing at the bottom of Fig. 230-10 would have to be either 10 or 12 ft above the balcony’s floor surface.

In part (C) of this section the Code says that service-drop or service-entrance conductors must not be mounted on or secured to a building wall directly beneath an elevated opening through which supplies or materials are moved into and out of the building. Such installations of conductors—say, beneath a high door to a barn loft—would obstruct access to the opening and present a hazard to personnel (Fig. 230-12).

Fig. 230-12. This violates the rule of the last paragraph of 230.9.

230.22. Insulation or Covering. In the past, the use of “covered”—not “insulated”—wire, such as TBWP (triple-braid weatherproof wire), resulted in quite a few tragic accidents, including a number of electrocutions. As a result, for many years now, only the use of insulated wire for ungrounded conductors was permitted with service conductors. For the 2002 NEC, however, Code-making panel (CMP) 4 has, with very little substantiation, gone ahead and again recognized the use of “covered” instead of insulated conductors for overhead service conductors. The panel did specifically mention the legitimacy of this application for medium voltage applications. There were no comments received during the comment period, either positive or negative, on this change.

The Exception recognizes the use of a bare grounded (neutral) conductor of a multiconductor cable. The exception only covers multiconductor cables, and therefore grounded neutral of open wiring must be insulated or covered just as the ungrounded conductors.

230.24. Clearances. There are four exceptions to the basic rule of part (A) that service-drop conductors must have at least an 8-ft (2.5-m) vertical clearance from the highest point of roofs over which they pass.

Exception No. 1 to the basic rule calling for 8-ft (2.5-m) clearance of service-drop conductors above a roof requires that clearance above a flat roof subject to pedestrian traffic or used for auto and/or truck traffic must observe the heights for clearance of drop conductors from the ground as given in part (B) of 230.24.

The intent of Exception No. 2 is that where the roof has a slope greater than 4 in. (100 mm) in 12 in. (300 mm), it is considered difficult to walk upon, and the height of conductors could then be less than 8 ft (2.5 m) from the highest point over which they pass but in no case less than 3 ft (900 mm) except as permitted in Exception No. 3. Figure 230-13 shows the rule. Exception No. 4 eliminates the need for maintaining the 8-ft minimum for 3 ft vertically in all directions where the final span attaches to the side of the building. This exception is particularly useful for a service drop hitting a building on a front corner above a porch roof below it. Without this provision the drop would have to attach at a great height or else a second pole would be required to redirect the drop so it missed the projected footprint of the porch roof. Figure 230-14 shows the conditions permitted by Exception No. 3 and Exception No. 4.

Part (B) covers service-drop clearance to ground, as shown in Fig. 230-15. The four dimensions of clearance from ground—10, 12, 15, and 18 ft (3.0, 3.7, 4.5, and 5.5 m)—are qualified by voltage levels and, for the 10-ft (3.0-m) mounting

Fig. 230-13. Service-drop conductors may have less than 8-ft (2.5-m) roof clearance. (Sec. 230.24.)

Fig. 230-14. Reduced clearance for service drop. (Sec. 230.24.)

height, by the phrase “only for service-drop cables.” These NE Code rules are generally in agreement with the National Electrical Safety Code. Where mast-type service risers are provided, the clearances in 230.24(B) will have to be considered by the installer.

Fig. 230-15. Service-drop clearance to ground. (Sec. 230.24.)

230.28. Service Masts as Supports. Figure 230-16 illustrates this rule. The last sentence of the rule is both important and controversial. It disallows all drops except the service drop from colocating on a service mast. No telephone drops, coaxial cable drops, or any other drop. This is an absolute prohibition and it applies no matter how stout the mast, no matter how well the mast is guyed, no matter how long the mast, and no matter what spacing would be provided between the service drop and other prospective drops from other utilities.

230.30. Insulation. This rule presents the requirement that service lateral conductors must be insulated. Although service-drop conductors were previously required to be insulated, CMP 4 has seen fit to reinstate permission to use “covered” overhead service conductors, as given in 230.22. No such permission is granted for underground service conductors.

The Exceptions to 230.30 and 230.41(A) clarify the use of aluminum, copper-clad aluminum, and bare copper conductors used as grounded conductors in service laterals and service-entrance conductors (Fig. 230-17).

For service lateral conductors (underground service), an individual grounded conductor (such as a grounded neutral) of aluminum or copper-clad aluminum without insulation or covering may not be used in a raceway underground. A bare copper neutral may be used—in a raceway, in a cable assembly,

Fig. 230-16. Service mast must provide adequate support for connecting drop conductors. (Sec. 230.28.)

or even directly buried in soil where local experience establishes that soil conditions do not attack copper.

The wording of part (4) of the Exception permits an aluminum grounded conductor of an underground service lateral to be without individual insulation or covering “when part of a cable assembly identified for underground use” where the cable is directly buried or run in a raceway. Of course, a lateral made up of individual insulated phase legs and an insulated neutral is acceptable in underground conduit or raceway (Fig. 230-18).

230.32. Protection Against Damage. Underground service lateral conductors—whether directly buried cables, conductors in metal conduit, or conductors in nonmetallic conduit—must comply with 300.5 for protection against physical damage. But WATCH OUT! Where conductors are buried at depths of 450 mm (18 in.) or more below grade, compliance with a special rule in Sec. 300.5 for service conductors is mandatory. As called for by Sec. 300.5(D)(3), the local inspector will always require that a warning ribbon be buried in the trench not less than a certain distance (i.e., 300 mm [12 in.]) above the buried service lateral or buried service entrance conductors (Fig. 230-19).

230.33. Spliced Conductors. Service conductors in the form of underground service laterals and all service entrance conductors are permitted to be spliced as long as the splicing method complies with the usual rules in the NEC for general wiring of comparable size and location. The NEC does not expressly cover splices in overhead service drops, but given the other rules as long as the splice meets industry standards for strain tolerance and workmanship, it would normally be permitted subject to the judgment of the authority having jurisdiction.

Fig. 230-17. 230.30 and 230.41 permit bare neutrals for service conductors. (Secs. 230.30 and 230.41.)

230.40. Number of Service-Entrance Conductor Sets. As a logical follow-up to the basic rule of 230.2, which requires that a single building or structure must be supplied “by only one service” (i.e., only one service drop or lateral), this rule calls for only one set of service-entrance (SE) conductors to be supplied by each service drop or lateral that is permitted for a building. Exception No. 1 covers a multiple-occupancy building (a two-family or multifamily building, a multitenant office building, or a store building, etc.). In such cases, a set of SE conductors for

Fig. 230-18. Underground bare aluminum grounded leg must always be in a cable assembly. (Sec. 230.30.)

Fig. 230-19. Protecting underground service conductors. (Sec. 230.32.)

each occupancy or for groups of occupancies is permitted to be tapped from a single drop or lateral (Fig. 230-20).

When a multiple-occupancy building has a separate set of SE conductors run to each occupancy, in order to comply with 230.70(A), the conductors should either be run on the outside of the building to each occupancy or, if run inside

Fig. 230-20. Service layouts must simultaneously satisfy 230.2, 230.40, 230.71, and all other NEC rules that are applicable. (Sec. 230.40.)

the building, be encased in 2 in. (50.8 mm) of concrete or masonry in accordance with 230.6. In either case the service equipment should be located “nearest to the entrance of the conductors inside the building,” and each occupant would have to have “access to his disconnecting means.”

Any desired number of sets of service-entrance conductors may be tapped from the service drop or lateral, or two or more subsets of service-entrance conductors may be tapped from a single set of main service conductors, as shown for the multiple-occupancy building in Fig. 230-20.

As written, there are no limitations on this permission comparable to the parallel allowance in 230.2(B)(1) that allows multiple services to supply individual occupancies where there is no common location available for a conventional service. That permission only operates by special permission. Part of the special permission process can and usually does allow a review of reciprocal labeling. In the case of 230.40 Exception No. 1, since there is only one service, 230.2(E) does not apply. For example, suppose a multitenant building has seven occupancies. This allowance can result in a group of six disconnects (that being the limit in any one location) in the vicinity of the service drop or lateral, and then a seventh set of service entrance conductors extended around the building (or through concrete) to the seventh occupancy. Since 230.2(E) does not apply, there is absolutely no requirement to post a sign or directory at the principal service location advising emergency personnel that opening the six disconnects at that location does not, in fact, disconnect the entire building. Given the cost of providing a master disconnect for a group of services, this is far from a purely academic concern. Some jurisdictions have placed limits on this allowance for that reason.

Exception No. 2 permits two to six disconnecting means to be supplied from a single service drop or lateral where each disconnect supplies a separate load (Fig. 230-21). Exception No. 2 recognizes the use of, say, six 400-A sets of service-entrance conductors to a single-occupancy or multiple-occupancy building in lieu of a single main 2500-A service. It recognizes the use of up to six subdivided loads extending from a single drop or lateral in a single-occupancy as well as multiple-occupancy building. Where single metering is required, doughnut-type CTs could be installed at the service drop.

The real importance of this rule is to eliminate the need for “paralleling” conductors of large-capacity services, as widely required by inspection authorities to satisfy previous editions of the NEC (Fig. 230-21). This same approach could be used in subdividing services into smaller load blocks to avoid the use of the equipment ground-fault circuit protection required by 230.95.

This rule can also facilitate expansion of an existing service. Where less than six sets of service-entrance conductors were used initially, one or more additional sets can be installed subsequently without completely replacing the original service. Of course, metering considerations will affect the layout.

But, the two to six disconnects (circuit breakers or fused switches) must be installed close together at one location and not spread out in a building. Since under this exception the disconnects are still grouped, the objection raised under Exception No. 1 does not apply.

Fig. 230-21. Tapping sets of service-entrance conductors from one drop (or lateral). (Sec. 230.40.)

Exception No. 3 recognizes tapping two sets of service conductors from a single drop or lateral at a dwelling unit to supply the dwelling and one other building. Exception No. 4 recognizes an additional set of service entrance conductors to supply the “common area” panel required by Sec. 210.25. And Exception No. 5 specifically recognizes supplying “other” equipment as indicated in 230.82(5) and 230.82(6).

230.41. Insulation of Service-Entrance Conductors. Except for use of a bare neutral, as permitted, all service-entrance conductors must be insulated and may not simply be “covered”—as discussed under 230.22. The wording used in part (3) of the Exception in 230.41 is slightly different from that described previously for 230.30. In this section, the reference is to “service-entrance conductors” instead of “service lateral conductors.” But here, a bare individual aluminum or copper-clad aluminum grounded conductor (grounded neutral or grounded phase leg) may be used in a raceway or a cable assembly or for direct burial where “identified” for direct burial.

Aluminum SE cable with a bare neutral may be used aboveground as an SE conductor. But, an aluminum SE cable with a bare neutral may be used underground only if it is “identified” for underground use in a raceway or directly buried. Conventional-style SE-U aluminum SE cable with a bare neutral is not “identified” for use underground but may be used, as the first sentence of 230.40 describes, for “service-entrance conductors entering or on the exterior of buildings or other structures.” In “SE-U,” the “U” stands for “unarmored” not “underground.”

230.42. Minimum Size and Rating. Sizing of service-entrance conductors involves the same type of step-by-step procedure as set forth for sizing feeders covered in Art. 220. A set of service-entrance conductors is sized just as if it were a feeder. In general, the service-entrance conductors must have a minimum ampacity—current-carrying capacity—selected in accordance with the ampacity tables and rules of 310.15, as well as the rules for “continuous loading” following part (A), here, that is sufficient to handle the total lighting and power load as calculated in accordance with Art. 220. Where the Code gives demand factors to use or allows the use of acceptable demand factors based on sound engineering determination of less than 100 percent demand requirement, the lighting and power loads may be modified.

According to the last sentence of 230.42(A), the “maximum allowable current” of busways used as service-entrance conductors must be taken to be the ampere value for which the busway has been listed or labeled. This is an “exception” to the basic rule that requires the ampacity of service-entrance conductors to be “determined from 310.15”—which does not give ampacities of busways. Parts (A)(1) and (A)(2) repeat the rule of 215.2(A) and its Exception. (See 215.2.)

From the analysis and calculations given in the feeder-circuit section, a total power and lighting load can be developed to use in sizing service-entrance conductors. Of course, where separate power and lighting services are used, the sizing procedure should be divided into two separate procedures.

When a total load has been established for the service-entrance conductors, the required current-carrying capacity is easily determined by dividing the total load in kilovoltamperes (or kilowatts with proper correction for power factor or the load) by the voltage of the service.

From the required current rating of conductors, the required size of conductors is determined. Sizing of the service neutral is the same as for feeders. Although suitably insulated conductors must be used for the phase conductors of service-entrance feeders—except as permitted for overhead conductors as described in 230.22—the NE Code does permit use of bare grounded conductors (such as neutrals) under the conditions covered in 230.30 and 230.41.

An extremely important element of service design is that of fault consideration. Service busway and other service conductor arrangements must be sized and designed to ensure safe application with the service disconnect and protection. That is, service conductors must be capable of withstanding the let-through thermal and magnetic stresses on a fault.

After calculating the required circuits for all the loads in the electrical system, the next step is to determine the minimum required size of service-entrance conductors to supply the entire connected load. The NE Code procedure for sizing SE conductors is the same as for sizing feeder conductors for the entire load—as set forth in 215.2(A).

Basically, the service “feeder” capacity must be not less than the sum of the loads on the branch circuits for the different applications.

The general lighting load is subject to demand factors from Table 220.12, which takes into account the fact that simultaneous operation of all branch-circuit loads, or even a large part of them, is highly unlikely. Thus, service or feeder capacity does not have to equal the connected load. The other provisions of Art. 220 are then factored in.

Reference to 230.79 in part (B) of 230.42 makes a 100-A service conductor ampacity a mandatory minimum if the system supplied is a one-family dwelling. And for all other occupancies where more than two 2-wire circuits are supplied, the minimum rating of the service conductors may not be less than 60 A. This reference is not intended to require the service conductors to always have an ampacity equal to the rating of the service disconnect(s). That is, for those installations described in 230.79, the service conductors must have the minimum ampacity required by 230.79. But for all other installations, the ampacity as established in accordance with 310.15 must be not less than the calculated load as determined in accordance with Art. 220, and the service conductors must be sized as required by 230.42(A), for continuous loading.

Another point of confusion is the wording here, “less than the rating of the service disconnecting means specified....” This does not mean that if you install a 400-A fused service switch for a 250-A load and with 250-A fuses, the service needs to be cabled for 400 A. It simply means that in the specified applications, the conductors must meet the minimums set in 230.79. The following table correlates the requirements for service lateral, service drop, and service entrance conductors in a single location.

230.43. Wiring Methods for 600 V, Nominal, or Less. The list of acceptable wiring methods for running service-entrance conductors does include flexible metal conduit (Greenfield) and liquidtight flexible metal conduit, but limits use of such raceways to a maximum length of 6 ft (1.8 m), and an equipment bonding conductor must be run with it. Although such raceways were prohibited at one time, effectively bonded flexible metal conduit and liquidtight flexible metal conduit in a length not over 6 ft (1.8 m) may be used as a raceway for service-entrance conductors (Fig. 230-22). A length of flex or liquidtight flex not longer than 6 ft (1.8 m)—in total—may be used as a service raceway, provided an equipment bonding conductor sized from Table 250.66 (and with a cross-sectional area at least 12½ percent of the csa of the largest service phase conductor for conductors larger than 1100 MCM copper or 1750 MCM aluminum) is used. This rule recognizes that the flexibility of such raceway is often needed or desirable in routing service-entrance conductors around obstructions in the path of connections between metering equipment and service-entrance switchboards, panelboards, or similar enclosures. The required equipment grounding conductor may be installed either inside or outside the flex, using acceptable fittings and termination techniques for the grounding conductor.

It should be noted that liquidtight flexible metal conduit is recognized as an acceptable service raceway, provided the bonding requirements given in 250.102 are satisfied. And, liquidtight flexible nonmetallic conduit—of any length—may be used as a service raceway containing service-entrance conductors.

230.44. Cable Trays. This section recognizes the use of a cable tray for the support of service-entrance conductors, provided the cable tray contains only service-entrance conductors. The exception permits service-entrance conductors in the same cable tray with other conductors, provided a “solid fixed barrier” is installed in the cable tray between the service conductors and the other, non-service conductors within the cable tray. In addition, a cable tray used under this exception must be marked “Service-Entrance Conductors” with permanently attached labels located so the routing of the service-entrance conductors may be easily traced from start to finish.

Fig. 230-22. These two flexible conduits may be used for service raceway. (Sec. 230.43.)

230.46. Unspliced Conductors. The wording in this section used to prohibit splicing of SE conductors, and that prohibition was followed by a number of exceptions. However, that long-standing rule was eliminated in the 1999 NEC. The Code now makes specific references to recognized methods, that is, “clamped or bolted connections.” Now splicing of SE conductors may be accomplished using the methods described by the applicable rules in 110.14, 300.13, and 300.15, which recognize use of the splicing methods previously described. (See Fig. 230-23.)

230.50. Protection of Open Conductors and Cables Against Damage—Aboveground. The wording in part (A) of this section no longer contains the “laundry list” of instances where physical protection would be required. Instead, it is now strictly up to the local inspector to determine where such protection must be provided. It seems reasonable to assume that those cases previously identified would be covered.

In parts (A)(1) through (5), the Code indicates specific methods that may be utilized to achieve the desired and required physical protection. In keeping with the broad discretion given by the wording used in this rule, part (A)(5) recognizes “other approved means.” Given the definition of the term “approved” in Art. 100, that wording essentially means “whatever the inspector will accept.” Since it is up to the inspector to decide when such protection is needed, it seems reasonable to grant the inspector the latitude to establish another method to ensure the required physical protection needed for service cables (Fig. 230-24).

Fig. 230-23. Permitted splices in service-entrance conductors. (Sec. 230.46.)

Fig. 230-24. Outdoor service raceway must be raintight and drained and SE cable must be protected. Although it is no longer specifically mentioned, it seems as if all those installations previously mentioned are still required to have physical protection. (Secs. 230.50 and 230.53.)

Part (B) exception allows use of type MI and MC cables for service-entrance or service lateral applications, without need for mounting at least 10 ft (3.0 m) above grade—provided they are not exposed to damage or are protected.

230.51. Mounting Supports. Service-entrance cable must be clamped to building surface by straps at intervals not over 30 in. (750 mm). And the cable must still be clamped within 12 in. (300 mm) of the service weather head and within 12 in. (300 mm) of cable connection to a raceway or enclosure.

Where the cable assembly is not listed for attachment to building surfaces, in part (B) the Code calls for such cables to be mounted on insulators that provide at least 2 in. of clearance from the building surface, and the insulators must be mounted no more than 15 ft apart. Similar requirements are given in part (C) for service conductors run as open individual conductors. Table 230.51(C) spells out the maximum spacing for the insulators and the minimum clearance from the building surface.

230.53. Raceways to Drain. Service-entrance conductors in conduit must be made raintight, using raintight raceway fittings, and must be equipped with a drain hole in the service ell at the bottom of the run or must be otherwise provided with a means of draining off condensation (Fig. 230-24). Use of a “pitch” equivalent to in. per foot, or 1 in. per 8 ft (10.4 mm per meter) will serve to satisfy the rule given by the last sentence for arranging conduit to drain where embedded in masonry.

230.54. Overhead Service Locations. When conduit or tubing is used for a service, the raceway must be provided with a service head (or weather head). Figure 230-25 shows details of a service-head installation. As covered in the Exception to part (B), service cable may be installed without a service head, provided it is bent to form a “gooseneck;” then tape the end with a “self-sealing water-resistant thermoplastic: that is, where no service head is used at the upper end of a service cable, the cable should be bent over so that the individual conductors leaving the cable will extend in a downward direction, and the end of the cable should be carefully taped and painted or sealed with water-resistant tape to exclude moisture.

Fig. 230-25. Location of service head minimizes entrance of rain. (Sec. 230.54.) The violation at the lower right can often be cured by twisting the service head so it points straight down, thereby avoiding the arrangement to the left that many consider unsightly.

Part (C) of this section requires that service heads be located above the service-drop attachment. Although this arrangement alone will not always prevent water from entering service raceways and equipment, such an arrangement will solve most of the water-entrance problems. An exception to this rule permits a service head to be located not more than 24 in. (600 mm) from the service-drop termination where it is found that it is impractical for the service head to be located above the service-drop termination. In such cases a mechanical connector is advisable at the lowest point in the drip loop to prevent siphoning. This exception will permit the Code-enforcing authority to handle hardship cases that may occur.

As covered by the wording in part (D), service cables shall be “held securely in place.” And each phase and neutral must be routed through an individual bushed opening in the service head to satisfy the basic rule in 230.54(E). But, the exception following this rule permits deviation from the one-phase one-bushed opening where the service conductors are in a jacketed multiconductor cable, as would be required in the case of a gooseneck.

The intent of part (G) is to require use of connections or conductor arrangements, both at the pole and at the service, so that water will not enter connections and siphon under head pressure into service raceways or equipment.

230.56. Service Conductor with the Higher Voltage-to-Ground. This Code rule presents the requirement that the “high” leg (the 208-V-to-ground leg) of a 240/120-V 3-phase, 4-wire delta system must be identified by marking to distinguish it from the other hot legs, which are only 120 V to ground. One method permitted is color-coding the so-called high leg orange. The rule recognizes “other means,” but any such “identification” must be provided “at each termination or junction point.” Clearly, the use of an overall orange-colored insulation will most easily satisfy this rule.

230.70. General. Parts (A)(1), (2), and (3) cover the place of installation of a service disconnect. The disconnecting means required for every set of service-entrance conductors must be located at a readily accessible point outside the building; or, where installed inside, nearest to the point at which the service conductors enter the building (Fig. 230-26). The service disconnect switch (or circuit breaker) is generally placed on the inside of the building as near as possible to the point at which the conductors come in.

And part (B) requires lettering or a sign on the disconnect(s) to identify it (them) as “Service Disconnect.” There are no exceptions to this marking requirement. Most panels with an integral main breaker come with an embossed or stamped marking “MAIN” next to that breaker, along with “SERVICE DISCONNECT” labels to be applied in the event the panel will be used as service equipment. These labels must be applied as appropriate, and nowhere else. A main breaker and a service disconnect are two different things. A building disconnect and a service disconnect are two different things. In many cases the function is obvious, but not all. This author recalls having been called by fire officials to a convenience store with a line up of six identically sized panels, all with “Main” breakers and no other designation. Two of those breakers were service disconnects and four were not, as definitively determined only after spending 20 min actually removing the dead fronts and tracing conductors.

Fig. 230-26. Service disconnect must open current for any conductors within building. (Sec. 230.70.)

Fortunately there was no emergency at the time; the department was engaged in a valuable exercise in advance preparation. While part (C) calls for the equipment to be “suitable” for use as service equipment, in all practicality this means that such disconnects must be listed and marked as being suitable for use as service equipment.

Although the Code does not set any maximum distance from the point of conductor entry to the service disconnect, various inspection agencies set maximum limits on this distance. For instance, service cable may not run within the building more than 18 in. (450 mm) from its point of entry to the point at which it enters the disconnect. Or, service conductors in conduit must enter the disconnect within 10 ft (3.0 m) of the point of entry. Or, as one agency requires, the disconnect must be within 10 ft (3.0 m) of the point of entry, but overcurrent protection must be provided for the conductors right at the point at which they emerge from the wall into the building. The concern is to minimize the very real and proven potential hazard of having unprotected service conductors within the building. Faults in such unprotected service conductors must burn themselves clear and such application has caused fires and fatalities. Check with your local inspection agency to find out what it intends to enforce. Such action should serve to prevent any surprises on the job.

Often shifting the point of entry, with somewhat longer conductors run outdoors, will solve a problem. In extreme cases a combination meter and overcurrent protection/disconnect can be installed, which then allows the conductors to run anywhere in the building. Many jobs have issues with obstructions such as oil tanks that need to be discussed with the inspector. Every cycle the Code-making panel receives, and rejects, proposals to set a specific allowable distance on the indoor length of service conductors. The panel intends that this remain a topic of negotiation and discussion between installers and inspectors with respect to the specific problems that arise in the field.

230.71. Maximum Number of Disconnects. Service-entrance conductors must be equipped with a readily accessible means of disconnecting the conductors from their source of supply. As stated in part (A), the disconnect means for each service and each set of SE conductors permitted by Sec. 230.2 and Sec. 230.40 Exception No. 1 and Exception No. 3, respectively, may consist of not more than six switches or six circuit breakers, in a common enclosure or grouped individual enclosures, located either “within sight of” and outside the building wall, or inside, as close as possible to the point at which the conductors enter the building. Figure 230-27 shows the basic application of that rule to a single set of SE conductors.

The last sentence in part (A) identifies a number of specific applications where the disconnecting means is not to be counted as “service disconnecting means” and applied against the maximum number of six disconnects permitted in the first two sentences. That is, when control power for a ground-fault protection system is tapped from the line side of the service disconnect means, the disconnect for the control power circuit is not counted as one of the six permitted disconnects for a service. A ground-fault-protected switch or circuit breaker supplying power to the building electrical system counts as one of the six permitted disconnects. But a disconnect supplying only the control-circuit power for a ground-fault protection system, installed as part of the listed equipment, does not count as one of the six service disconnects. The same idea applies to all other specifically identified equipment disconnects that are generally located at services, including surge protective devices and power monitoring equipment.

The rule of this section correlates “number of disconnects” with 230.2 and 230.40, which permit a separate set of SE conductors to be run to each occupancy (or group of occupancies) in a multiple-occupancy building, as follows:

230.2permits more than one “service” to a building—that is, more than one service drop or lateral—under the conditions set forth. As set forth in the first sentence of 230.40 each such “service” must supply only one set of SE conductors in a building that is a single-occupancy (one-tenant) building, and each set of SE conductors may supply up to six SE disconnects grouped together at one location—in the same panel or switchboard or in grouped individual enclosures. If the grouped disconnects for one set of SE conductors are not at the same location as the grouped disconnects for one or more other sets of SE conductors, for those situations described and permitted in 230.2, then a “plaque or directory” must be placed at each service-disconnect grouping to tell where

Fig. 230-27. The three basic ways to provide service disconnect means. (Sec. 230.71.)

the other group (or groups) of disconnects are located and what loads each group of disconnects serves.

Exception No. 1 to 230.40 says that a single service drop or lateral may supply more than one set of SE conductors for a multiple-occupancy building. Then at the load end of each of the sets of SE conductors, in an individual occupancy or adjacent to a group of occupancy units (apartments, office, stores), up to six SE disconnects may be supplied by each set of SE conductors. And Exception No. 3 recognizes two sets of SE conductors at a dwelling unit to supply the dwelling and one other separate “structure.”

The first sentence of part (A) to 230.71 ties directly into 230.40, Exception No. 1. It is the intent of this basic rule that, where a multiple-occupancy building is provided with more than one set of SE conductors tapped from a drop or lateral, each set of those SE conductors may have up to six switches or circuit breakers to serve as the service disconnect means for that set of SE conductors. The rule does recognize that six disconnects for each set of SE conductors at a multiple-occupancy building with, say, 10 sets of SE conductors tapped from a drop or lateral does result in a total of 6 × 10, or 60, disconnect devices for completely isolating the building’s electrical system from the utility supply. 230.72(B) also recognizes use of up to six disconnects for each of the “separate” services for fire pumps, emergency lighting, and so on, which are recognized in 230.2 as being separate services for specific purposes. And if service is provided for different classes of service, six disconnects could be provided for each class of service to each occupancy, resulting in 120 disconnects!

Although the basic rule of 230.40 specifies that only one set of SE conductors may be tapped from a single drop for a building with single occupancy, Exception No. 2 to 230.40 recognizes that a separate set of SE conductors may be run from a single service drop or lateral to each of up to six service disconnects mounted in separate enclosures at one location, constituting the disconnect means for a single service to a single-occupancy building.

For any type of occupancy, a panel containing up to six switches or circuit breakers may be used as service equipment with the enclosed six or fewer breakers comprising the service disconnecting means under a special exception (408.36 Exception No. 1). A panel used as service equipment for renovation of an existing service in an individual residential occupancy (but not for new installations) may have up to six main breakers or fused switches under 408.36 Exception No. 3. A panel meeting the old 42-circuit limitation for the former lighting and appliance branch-circuit panelboards may have up to two main breakers (or sets of fuses), per 408.36 Exception No. 2. However, a panel used without these limitations and used as service equipment for new buildings of any type must have not more than a single main device—with its rating not greater than the panel bus rating. See 408.36.

The first sentence of 230.71(A) and that of 230.72(A) note that from one to six switches (or circuit breakers) may serve as the service disconnecting means for each class of service for a building. For example, if a single-occupancy building has a 3-phase service and a separate single-phase service, each such service may have up to six disconnects (Fig. 230-28). Where the two sets of service equipment are not located adjacent to each other, a plaque or directory must be installed at each service-equipment location indicating where the other service equipment is—as required by 230.2(E).

230.71(B) notes that single-pole switches or circuit breakers equipped with handle ties may be used in groups as single disconnects for multiwire circuits, simultaneously providing overcurrent protection for the service (Fig. 230-29). Multipole switches and circuit breakers may also be used as single disconnects. The requirements of the Code are satisfied if all the service-entrance conductors can be disconnected with no more than six operations of the hand—regardless of whether each hand motion operates a single-pole unit, a multipole unit, or a group of single-pole units with “handle ties” or a “master handle” controlled by a single hand motion. Of course, a single main device for service disconnect and overcurrent protection—such as a main CB or fused switch—gives better protection to the service conductors.

The FPN to this section refers to 408.36 Exceptions 1 and 3, which vary from the individual protection requirements that now apply to panelboards generally,

Fig. 230-28. Each separate service may have up to six disconnect devices. (Sec. 230.71.)

Fig. 230-29. This arrangement constitutes six disconnects. (Sec. 230.71.)

and as discussed previously. The reference to 430.95 is intended to point out the limitations associated with installations where the service equipment is within a motor control center. For such installations, the rule of 430.95 mandates the use of a single main disconnect.

230.72. Grouping of Disconnects. The basic rule of part (A) requires that for a service disconnect arrangement of more than one disconnect—such as where two to six disconnect switches or CBs are used, as permitted by 230.71(A)—all the disconnects making up the service equipment “for each service” must be grouped and not spread out at different locations. The basic idea is that anyone operating the two to six disconnects must be able to do it while standing at one location. Service conductors must be able to be readily disconnected from all loads at one place. And each of the individual disconnects must have lettering or a sign to tell what load it supplies (Fig. 230-30).

Fig. 230-30. Two to six disconnect switches or CBs must be grouped and identified. (Sec. 230.72.)

This rule makes clear that the two to six service disconnects that are permitted by 230.71(A) for each “service” or for “each set of SE conductors” at a multiple-occupancy building must be grouped. But, where permitted by 230.2, the individual groups of two to six breakers or switches do not have to be together, and if they are not together, a sign at each location must tell where the other service disconnects are. (See 230.2.) Each grouping of two to six disconnects may be within a unit occupancy—such as an apartment—of the building.

The exception to part (A) permits (Note: It permits, it does not require, but read the next paragraphs) one of the two to six service disconnects to be located remote from the other disconnecting means that are grouped in accordance with the basic rule—PROVIDED THAT the remote disconnect is used only to supply a water pump that is intended to provide fire protection. In a residence or other building that gets its water supply from a well, a spring, or a lake, the use of a remote disconnect for the water pump will afford improved reliability of the water supply for fire suppression in the event that fire or other faults disable the normal service equipment. And it will distinguish the water-pump disconnect from the other normal service disconnects, minimizing the chance that firefighters will unknowingly open the pump circuit when they routinely open service disconnects during a fire. This exception ties into the rule of 230.72(B), which requires (not simply permits) remote installation of a fire-pump disconnect switch that is required to be tapped ahead of the one to six switches or CBs that constitute the normal service disconnecting means [see 230.82(5)]. The exception provides remote installation of a normal service disconnect when it is used for the same purpose (water pump used for fire fighting) as the emergency service disconnect (fire pump) covered in 230.72(B). In both cases, remote installation of the pump disconnect isolates the critically important pump circuit from interruption or shutdown due to fire, arcing-fault burndown, or any other fault that might knock out the main (normal) service disconnects.

A wide variety of layouts can be made to satisfy the Code permission for remote installation of a disconnect switch or CB service as a normal service disconnect (one of a maximum of six) supplying a water pump. Figure 230-31 shows three typical arrangements that would basically provide the isolated fire-pump disconnect.

Part (B), as noted above, makes it mandatory to install emergency disconnect devices where they would not be disabled or affected by any fault or violent

Fig. 230-31. Rule permits remote installation of one of two-to-six service disconnects to protect fire-pump circuits (typical layouts). (Sec. 230.72.)

electrical failure in the normal service equipment (Fig. 230-32). Figure 230-33 shows a service disconnect for emergency and exit lighting installed very close to the normal service switchboard. An equipment burndown or fire near the main switchboard might knock out the emergency circuit. And the tap for the switch, which is made in the switchboard ahead of the service main, is particularly susceptible to being opened by an arcing failure in the board. The switch should be 10 or 15 ft (3.0 or 4.5 m) away from the board. And because the switchboard is fed from an outdoor transformer-mat layout directly outside the building, the tap to the safety switch would have greater reliability if it was made from the transformer secondary terminals rather than from the switchboard service terminals. Although the rule sets no specific distance of separation, remote locating of emergency disconnects is a mandatory Code rule.

Fig. 230-32. Emergency service disconnects must be isolated from faults in normal SE equipment. (Sec. 230.72.)

Fig. 230-33. Emergency disconnect close to service switchboard and fed by tap from it could readily be disabled by fault in board. (Sec. 230.72.)

In part (B), the phrase “permitted by 230.2” makes clear that each separate service permitted for fire pumps or for either legally required or optional standby service may be equipped with up to six disconnects in the same way as the normal service—or any service—may have up to six SE disconnects. And the disconnect or disconnects for a fire-pump or standby services must be remote from the normal service disconnects, as shown in Fig. 230-32.

Part (C) applies to applications of service disconnect for multiple-occupancy buildings—such as apartment houses, condominiums, town houses, office buildings, and shopping centers. Part (C) requires that the disconnect means for each occupant in a multiple-occupancy building be accessible to each occupant. For instance, for the occupant of an apartment in an apartment house, the disconnect means for deenergizing the circuits in the apartment must be in the apartment (such as a panel), in an accessible place in the hall, or in a place in the basement or outdoors where it can be reached.

As covered by the exception to part (C), the access for each occupant as required by paragraph (C) would be modified where the building was under the management of a building superintendent or the equivalent and where electrical service and maintenance were furnished. In such a case, the disconnect means for more than one occupancy may be accessible only to authorized personnel. Figure 230-34 summarizes the way the grouping requirements are typically applied to multiple occupancy applications.

Fig. 230-34. These groupings of service disconnects represent good and acceptable practice that has been followed widely. (Sec. 230.72.)

230.75. Disconnection of Grounded Conductor. In this section the other means for disconnecting the grounded conductor from the interior wiring may be a screw or bolted lug on the neutral terminal block. The grounded conductor must not be run straight through the service equipment enclosure with no means of disconnection.

230.76. Manually or Power Operable. Any switch or CB used for service disconnect must be manually operable. In addition to manual operation, the switch may have provision for electrical operation—such as for remote control of the switch, provided it can be manually operated to the open or OFF position.

Code wording clearly indicates that an electrically operated breaker with a mechanical trip button which will open the breaker even if the supply power is dead is suitable for use as a service disconnect. The manually operated trip button ensures that the breaker “can be opened by hand.” To provide manual closing of electrically operated circuit breakers, manufacturers provide emergency manual handles as standard accessories. Thus such breaker mechanisms can be both closed and opened manually if operating power is not available, which fully satisfies this rule (Fig. 230-35).

Local requirements on the use of electrically operated service disconnects should be considered in selecting such devices.

Fig. 230-35. Manual operation of any service switch is required. (Sec. 230.76.)

230.79. Rating of Disconnect. Aside from the limited conditions covered in parts (A) and (B), this section requires that service equipment (in general) shall have a rating not less than 60 A, applicable to both fusible and CB equipment.

Part (C) requires 100-A minimum rating of a single switch or CB used in the service disconnect for any “one-family dwelling.” It should be noted that the rule applies to one-family houses only, because of the definition of “one-family dwelling” as given in Art. 100. It does not apply to apartments or similar dwelling units that are in two-family or multifamily dwellings. These rules and the requirements in 230.42(B) must be carefully correlated. Review the discussion at 230.42(B) in this chapter for more information on this topic.

Even if the demand on a total connected load, as calculated from 220.40 through 220.61 or any of the applicable optional calculations permitted by part (C) of Art. 220, is less than 10 kVA, a 100-A service disconnect, as well as 100-A rated service-entrance conductors [230.42(B)], must be used.

If a 100-A service is used, the demand load may be as high as 24 kVA. By using the optional service calculations of Table 220.82, a 24-kVA demand load is obtained from a connected load of as much as 45 kVA, depending on how the load is configured. This shows the effect of diversity on large-capacity installations.

230.80. Combined Rating of Disconnects. Figure 230-36 shows an application of this rule, based on determining what rating of a single disconnect would be required if a single disconnect were used instead of multiple ones. It should be noted that the sum of ratings above 400 A does comply with the rule of this section and with Exception No. 3 of 230.90(A), even though the 400-A service-entrance conductors could be heavily overloaded. Exception No. 3 exempts this type of layout from the need to protect the conductors at their rated ampacity, as required in the basic rule of 230.90. The Code assumes that the 400-A rating of the service-entrance conductors was carefully calculated from Art. 220 to be adequate for the maximum sum of the demand loads fed by the five disconnects shown in the layout.

Fig. 230-36. Multiple disconnects must have their sum of ratings at least equal to the minimum rating of a single disconnect. (Sec. 230.80.)

230.82. Equipment Connected to the Supply Side of Service Disconnect. Cable limiters, fuses or CBs away from the building, high-impedance shunt circuits (such as potential coils of meters, etc.), supply conductors for time switches, surge-protective capacitors, instrument transformers, lightning arresters and circuits for emergency systems, fire-pump equipment, and fire and sprinkler alarms may be connected on the supply side of the disconnecting means. Emergency-lighting circuits, surge-protective capacitors, and fire-alarm and other protective signaling circuits, when placed ahead of the regular service disconnecting means, must have separate disconnects and overcurrent protection.

Part (1) of the rule prohibiting equipment connections on the line side of the service disconnect permits “cable limiters or other current-limiting devices” to be so connected.

Cable limiters are used to provide protection for individual conductors that are used in parallel (in multiple) to make up one phase leg of a high-capacity circuit, such as service conductors. A cable limiter is a cable connection device that contains a fusible element rated to protect the conductor to which it is connected.

As indicated in Part (2), meters and meter sockets can be connected on the supply side of the service disconnecting means and overcurrent protective devices if the meters are connected to service not in excess of 600 V where the grounded conductor bonds the meter equipment cases and enclosures to the grounding electrode. (See Fig. 230-37.)

Fig. 230-37. A “meter” may be connected on the supply side of the service disconnect [Sec. 230.82(2).]

Part (3) allows meter disconnect switches ahead of a service disconnect, provided they have a suitable load rating and fault-current interrupting capability. WARNING: Many bypass meter sockets contain a switch that is used to maintain continuity across the meter jaws when the meter is removed. Utility personnel will close this switch, remove the meter for service, replace the meter, and then open the switch. It is of paramount importance that this switch never be opened or closed under load. The switches are unsuitable for load switching, only to maintain continuity.

Electrical utilities are requiring meter disconnect switches with ever-increasing frequency for the safety of their metering departments, so they can work the metering equipment cold. For this reason they are an important safety enhancement. From a code enforcement perspective, they present some challenges and opportunities for conflict in the field, especially since they are now fully load-break rated. If such a fuse were fused to contribute to the fault duty of the switch, the switch would arguably contain all the elements of service equipment as defined in the NEC.

This immediately raises the question, did the location of the service disconnect just move? The orderly application of a plethora of NEC rules depend on common agreement on exactly what device constitutes the service disconnecting means for any premises wiring application. If the service disconnect, in effect, relocates, then the number of conductors to the building changes because the neutral cannot be used for equipment grounding, the permitted point of connecting the grounding electrode conductor changes, and on and on. Until this is clarified, be aware that there has never been any intent to change the intended service disconnect locations, and this switch is only for the use of utility employees, who may well lock the switch in the “ON” position except while it is under active use. However, since it may also meet the Art. 100 definition, it would be wise to review the plans and the utility site policies with the inspector so there will not be surprises at the time of the final service inspection.

Part (4) recognizes the connection of current and voltage transformers, high-impedance shunts, and surge protection on the line side of the service disconnecting means, but only where such devices are listed for such application. Where supply conductors are installed as service conductors, load management devices, fire alarm and suppression equipment, and standby power systems are permitted to be connected on the line side of the service disconnecting means by part (5). This part allows transfer switches ahead of the service disconnecting means in accordance with the limitations spelled out in 700.6, 701.7, and 702.6. Transfer switches are also available with ratings for service disconnecting means, in which case they will be marked accordingly.

As permitted by part (6), an electric power production source that is auxiliary or supplemental to the normal utility service to a premises may be connected to the supply (incoming) side of the normal service disconnecting means. This part of the rule permits connection of a solar photovoltaic system, fuel cells, or interconnected power sources into the electrical supply for a building or other premises, to operate as a parallel power supply.

Where properly protected and provided with suitable disconnects, control circuits for power-operated service disconnects may be connected ahead of the service disconnect as recognized by subpart (7). And part (8) recognizes that control power for a ground-fault protection system may be tapped from the supply side of the service disconnecting means. Where a control circuit for a ground-fault system is tapped ahead of the service main and “installed as part of listed equipment,” suitable overcurrent protection and a disconnect must be provided for the control-power circuit.

230.90. Where Required. The intent in paragraph (A) is to ensure that the over-current protection required in the service-entrance equipment protects the service-entrance conductors from “overload.” It is obvious that these overcurrent devices cannot provide “fault” protection for the service-entrance conductors if the fault occurs in the service-entrance conductors (which are on the line side of the service overcurrent devices), but they can protect the conductors from overload where so selected as to have proper rating. Conductors on the load side of the service equipment are considered as feeders or branch circuits and are required by the Code to be protected as described in Arts. 210, 215, and 240.

Part (A) states that the term “set” of fuses means all the fuses required to protect all the ungrounded service-entrance conductors in a given circuit.

Each ungrounded service-entrance conductor must be protected by an over-current device in series with the conductor (Fig. 230-38). The overcurrent device must have a rating or setting not higher than the allowable current capacity of the conductor, with the exceptions noted.

Fig. 230-38. Single main service protection must not exceed conductor ampacity (or may be next higher rated device above conductor ampacity). (Sec. 230.90.)

The rule of Exception No. 1 says that if the service supplies one motor in addition to other load (such as lighting and heating), the overcurrent device may be rated or set in accordance with the required protection for a branch circuit supplying the one motor (430.52) plus the other load, as shown in Fig. 230-39. Use of 175-A fuses where the calculation calls for 170-A conforms to Exception No. 2 of 230.90—next higher standard rating of fuse (240.6). For motor branch circuits and feeders, Arts. 220 and 430 permit the use of overcurrent devices having ratings or settings higher than the capacities of the conductors. Article 230 makes similar provisions for services where the service supplies a motor load or a combination load of both motors and other loads.

If the service supplies two or more motors as well as other load, then the overcurrent protection must be rated in accordance with the required protection for a feeder supplying several motors plus the other load (430.63). Or if the service supplies only a multimotor load (with no other load fed), then 430.62 sets the maximum permitted rating of overcurrent protection.

Exception No. 3. Not more than six CBs or six sets of fuses may serve as over-current protection for the service-entrance conductors even though the sum of the ratings of the overcurrent devices is in excess of the ampacity of the service

Fig. 230-39. Service protection for lighting plus motor load. (Sec. 230.90.)

conductors supplying the devices—as illustrated in Fig. 230-40. The grouping of single-pole CBs as multipole devices, as permitted for disconnect means, may also apply to overcurrent protection.

Exception No. 3 ties into 230.80. Service conductors are sized for the total maximum demand load—applying permitted demand factors from Art. 220. Then each of the two to six feeders fed by the SE conductors is also sized from Art. 220 based on the load fed by each feeder. When those feeders are given overcurrent protection in accordance with their ampacities, it is frequently found that the sum of those overcurrent devices is greater than the ampacity of the SE conductors, which were sized by applying the applicable demand factors to the total connected load of all the feeders. Exception No. 3 recognizes that possibility as acceptable even though it departs from the rule in the first

Fig. 230-40. With six subdivisions of protection, conductors could be overloaded. (Sec. 230.90.)

sentence of 230.90(A). The assumption is that if calculation of demand load for the SE conductors is correctly made, there will be no overloading of those conductors because the diversity of feeder loads (some loads “on,” some “off”) will be adequate to limit load on the SE conductors.

Assume that the load of a building computed in accordance with Art. 220 is 255 A. Under 240.4(B), 300-A fuses or a 300-A CB may be considered as the proper-size overcurrent protection for service conductors rated between 255 and 300 A if a single service disconnect is used.

If the load is separated in such a manner that six 70-A CBs could be used instead of a single service disconnecting means, total rating of the CBs would be greater than the ampacity of the service-entrance conductors. And that would be acceptable.

Exception No. 4 to 230.90(A) is shown in Fig. 230-41 and is intended to prevent opening of the fire-pump circuit on any overload up to and including stalling or even seizing of the pump motor. Because the conductors are “outside the building,” operating overload is no hazard; and, under fire conditions, the pump must have no prohibition on its operation. It is better to lose the motor than attempt to protect it against overload when it is needed.

Exception No. 5 specifically recognizes the use of conductors in accordance with 310.15(B)(6). There the Code considers the conductor sizes in Table 310.15(B)(6) to be adequately protected by the value of OC protection indicated.

230.95. Ground-Fault Protection of Equipment. Fuses and CBs, applied as described in the previous section on “Overcurrent Protection,” are sized to protect conductors in accordance with their current-carrying capacities. The function of a fuse or CB is to open the circuit if current exceeds the rating of the protective device. This excessive current might be caused by operating overload,

Fig. 230-41. (Sec. 230.90.)

by a ground fault, or by a short circuit. Thus, a 1000-A fuse will blow if current in excess of that value flows over the circuit. It will blow early on heavy over-current and later on low overcurrents. But it will blow, and the circuit and equipment will be protected against the damage of the overcurrent. But, there is another type of fault condition which is very common in grounded systems and will not be cleared by conventional overcurrent devices. That is the phase-to-ground fault (usually arcing) which has a current value less than the rating of the overcurrent device.

On any high-capacity feeder, a line-to-ground fault (i.e., a fault from a phase conductor to a conduit, to a junction box, or to some other metallic equipment enclosure) can, and frequently does, draw current of a value less than the rating or setting of the circuit protective device. For instance, a 500-A ground fault on a 2000-A protective device which has only a 1200-A load will not be cleared by the device. If such a fault is a “bolted” line-to-ground fault, a highly unlikely fault, there will be a certain amount of heat generated by the I²R effect of the current; but this will usually not be dangerous, and such fault current will merely register as additional operating load, with wasted energy (wattage) in the system. Further, such bolted faults usually draw large values of current, particularly if the equipment grounding system has been installed correctly, and the result will be a trip in very short order. But, bolted phase-to-ground faults are very rare. The usual phase-to-ground fault exists as an intermittent or arcing fault, and an arcing fault of the same current rating as the essentially harmless bolted fault can be fantastically destructive because of the intense heat of the arc.

Of course, any ground-fault current (bolted or arcing) above the rating or setting of the circuit protective device will normally be cleared by the device. But, even where the protective device eventually operates, in the case of a heavy ground-fault current which adds to the normal circuit load current to produce a total current in excess of the rating of the normal circuit protective device (fuse or CB), the time delay of the device may be minutes or even hours—more than enough time for the arcing-fault current to burn out conduit and enclosures, acting just like a torch, and even propagating flame to create a fire hazard.

In the interests of safety, definitive engineering design must account for protection against high-impedance ground faults, as is required by this Code rule. Phase overcurrent protective devices are normally limited in their effectiveness because (1) they must have a time delay and a setting somewhat higher than full load to ride through normal inrushes, and (2) they are unable to distinguish between normal currents and low-magnitude fault currents, which, when combined, may be less than the trip rating of the overcurrent protective device.

Dangerous temperatures and magnetic forces are proportional to current for overloads and short circuits; therefore, overcurrent protective devices usually are adequate to protect against such faults. However, the temperatures of arcing faults are, generally, independent of current magnitude; and arcs of great and extensive destructive capability can be sustained by currents not exceeding the overcurrent device settings. Other means of protection are therefore necessary. A ground-detection device, which “sees” only ground-fault current, is coupled to an automatic switching device to open all three phases when a line-to-ground fault exists on the circuit. Such protective systems are readily available in listed configurations from electrical equipment manufacturers, which eases compliance with this rule. Careful attention to manufacturer’s installation instructions is mandatory to ensure proper operation and the desired level of protection.

230.95 requires ground-fault protection of equipment (GFPE) to be provided for each service disconnecting means rated 1000 A or more in a solidly grounded-wye electrical service that operates with its ungrounded legs at more than 150 V to ground. Note that this applies to the rating of the disconnect, not to the rating of the overcurrent devices or to the capacity of the service-entrance conductors.

The wording of the first sentence of this section makes clear that service GFPE (ground-fault protection of equipment) is required under specific conditions: only for grounded-wye systems that have voltage over 150 V to ground and less than 600 V phase-to-phase. In effect, that means the rule applies only to 480/277-V grounded-wye and not to 120/208-V systems or any other commonly used systems (Fig. 230-42). Recent recognition of the 600Y/340-V distribution systems—used in Canada—would subject any system so rated to the rule of Sec. 230.95. And, each disconnect rated 1000 A, or more, must be provided with equipment GFPE. GFPE is not required on any systems operating over 600 V phase-to-phase. The reason for this voltage parameter is that on ac systems an arc naturally self-extinguishes at every current zero, 120 times per second. The likelihood of damage is directly related to the likelihood of the arc restriking, and that is related to the peak voltage in the system. Testing has shown this is a major problem above about 375 V. Since the peak voltage of a 120-V-to-ground system is 170 V (120 × ), these systems aren’t so much of a

Fig. 230-42. Service ground-fault protection is mandatory. (Sec. 230.95.)

problem. However, on 480Y/277-V systems, the peak voltage is about 390 V, easily high enough to keep the arc in business.

The second paragraph clearly indicates that the “rating” to be considered is the rating of the largest fuse a switch can accommodate, or the longtime trip rating of a nonadjustable CB, or the maximum “setting” for adjustable-trip CBs. If the fusible switch, nonadjustable CB, or adjustable-trip CB used as the service disconnect is rated at, or can be set at, 1000 A or more, then ground-fault protection is required.

In a typical GFPE hookup, as shown in Fig. 230-43, part (A) of the section specifies that a ground-fault current of 1200 A or more must cause the disconnect to open all ungrounded conductors. Thus the maximum GF pick-up setting permitted is 1200 A, although it may be set lower.

Fig. 230-43. GFPE is required for each disconnect rated 1000 A or more, but not for a fire-pump disconnect. (Sec. 230.95.)

With a GFPE system, at the service entrance a ground fault anywhere in the system is immediately sensed in the ground-relay system, but its action to open the circuit usually is delayed to allow some normal overcurrent device near the point of fault to open if it can. As a practical procedure, such time delay is designed to be only a few cycles or seconds, depending on the voltage of the circuit, the time-current characteristics of the overcurrent devices in the system, and the location of the ground-fault relay in the distribution system. Should any of the conventional short-circuit overcurrent protective devices fail to operate in the time predetermined to clear the circuit, and if the fault continues, the ground-fault protective relays will open the circuit. This provides added overcurrent protection not available by any other means.

The rule requiring GFPE for any service disconnect rated 1000 A or more (on 480/277-V or 600/347-V services) specifies a maximum time delay of 1 s for ground-fault currents of 3000 A or more (Fig. 230-44).

Fig. 230-44. The rule specifies maximum energy let-through for GFPE operation. [Sec. 230.95(A).]

The maximum permitted setting of a service GFPE hookup is 1200 A, but the time-current trip characteristic of the relay must ensure opening of the disconnect in not more than 1 s for any ground-fault current of 3000 A or more. This change in the Code was made to establish a specific level of protection under GFPE by setting a maximum limit on I²t of fault energy.

The reasoning behind this change was explained as follows:

The amount of damage done by an arcing fault is directly proportional to the time it is allowed to burn. Commercially available GFPE systems can easily meet the 1-s limit. Some users are requesting time delays up to 60 s so all downstream overcurrent devices can have plenty of time to trip thermally before the GFP on the main disconnect trips. However, an arcing fault lasting 60 s can virtually destroy a service equipment installation. Coordination with downstream overcurrent devices can and should be achieved by adding GFPE on feeder circuits where needed. The Code should require a reasonable time limit for GFP. Now, 3000 A is 250 percent of 1200 A, and 250 percent of setting is a calibrating point specified in ANSI 37.17. Specifying a maximum time delay starting at this current value will allow either flat or inverse time-delay characteristics for ground-fault relays with approximately the same level of protection.

Selective coordination between GFPE and conventional protective devices (fuses and CBs) on service and feeder circuits is now a very clear and specific task as a result of rewording of 230.95(A) that calls for a maximum time delay of 1 s at any ground-fault current value of 3000 A or more.

For applying the rule of 230.95, the rating of any service disconnect means shall be determined, as shown in Fig. 230-45.

Fig. 230-45. Determining rating of service disconnect for GFPE rule. (Sec. 230.95.)

Because the rule on required service GFPE applies to the rating of each service disconnect, there are many instances where GFPE would be required if a single service main disconnect is used but not if the service subdivision option of 230.71(A) is taken, as shown in Fig. 230-46.

Fig. 230-46. Subdivision option on disconnects affects GFPE rule. (Sec. 230.95.)

By the Exception to 230.95, continuous industrial process operations are exempted from the GFPE rules of parts (A), (B), and (C) where the electrical system is under the supervision of qualified persons who will effect orderly shutdown of the system and thereby avoid hazards, greater than ground fault itself, that would result from the nonorderly, automatic interruption that GFPE would produce in the supply to such critical continuous operations. The Exception excludes GFPE requirements where a nonorderly shutdown will introduce additional or increased hazards. The idea behind that is to provide maximum protection against service outage of such industrial processes. With highly trained personnel at such locations, design and maintenance of the electrical system can often accomplish safety objectives more readily without GFPE on the service. Electrical design can account for any danger to personnel resulting from loss of process power versus damage to electrical equipment.

The former Exception No. 2 at this location excluded fire-pump service disconnects from the basic rule that requires ground-fault protection on any service disconnect rated 1000 A or more on a grounded-wye 600/347-V or 480/277-V system. This exception has been deleted, not because it is a bad idea, but because the fire pump article now has a clear statement [at 695.6(H)] that forbids the use of GFPE on a fire pump circuit. Since per 90.3 a provision in Chap. 6 (or 5 or 7) automatically supersedes a contrary provision in Chaps. 1 through 4, an exception to the same effect here is a waste of space. There are good reasons for the prohibition.

Because fire pumps are required by 230.90, Exception No. 4, to have over-current protection devices large enough to permit locked-rotor current of the pump motor to flow without interruption, larger fire pumps (100 hp and more) would have disconnects rated 1000 A or more. Without Exception No. 2, those fire-pump disconnects would be subject to the basic rule and would have to be equipped with ground-fault protection. But GFP on any fire pump is objectionable on the same basis that 230.90, Exception No. 4, wants nothing less than protection rated for locked rotor. The intent is to give the pump motor every chance to operate when it functions during a fire, to prevent opening of the motor circuit or any overload up to and including stalling or seizing of the shaft or bearings. For the same reason, 430.31 exempts fire pumps from the need for overload protection, and 430.72(C) Exception requires overcurrent protection to be omitted from the control circuit of a starter for a fire pump.

Important considerations are given in fine-print notes in this section. Obviously, the selection of ground-fault equipment for a given installation merits a detailed study. The option of subdividing services discussed under six service entrances from one lateral [230.2(A)(1)] should be evaluated. A 4000-A service, for example, could be divided using five 800-A disconnecting means, and in such cases GFPE would not be required.

One very important note in 230.95 warns about potential desensitizing of ground-fault sensing hookups when an emergency generator and transfer switch are provided in conjunction with the normal service to a building. The note applies to those cases where a solid neutral connection from the normal service is made to the neutral of the generator through a 3-pole transfer switch. With the neutral grounded at the normal service and the neutral bonded to the generator frame, ground-fault current on the load side of the transfer switch can return over two paths, one of which will escape detection by the GFPE sensor, as shown in Fig. 230-47. Such a hookup can also cause nuisance tripping of the GFPE due to normal neutral current. Under normal (nonfaulted) conditions, neutral current due to normal load unbalance on the phase legs can divide at common neutral connection in transfer switch, with some current flowing toward the generator and returning to the service main on the conduit—indicating falsely that a ground fault exists and causing nuisance tripping of GFPE. The note points out that “means or devices” (such as a 4-pole, neutral-switched transfer switch) “may be needed” to ensure proper, effective operation of the GFPE hookup (Fig. 230-48).

Very Important!

Because of so many reports of improper and/or unsafe operation (or failure to operate) of ground-fault protective hookups, part (C) of 230.95 requires (a mandatory rule) that every GFPE hookup be “performance tested when first installed.” And the testing MUST be done on the job site! Factory testing of a GFPE system does not satisfy this Code rule. This rule requires that such testing be done according to “instructions . . . provided with the equipment.” A written record must be made of the test and must be available to the inspection authority.

Figure 230-49 shows two basic types of GFPE hookup used at service entrances.

Fig. 230-47. Improper operation of GFPE can result from emergency system transfer switch. (Sec. 230.95.)

Fig. 230-48. Four-pole transfer switch is one way to avoid desensitizing GFPE. (Sec. 230.95.)

Fig. 230-49. Types of ground-fault detection that may be selected for use at services. (Sec. 230.95.)

230.200. General (Services Exceeding 600 V, Nominal). The rules on medium-voltage services given in the provisions of Art. 230 apply only to equipment on the load side of the “service-point.” Because there has been so much controversy over identifying what is and what is not “service” equipment in the many complicated layouts of outdoor high-voltage circuits and transformers, the definition in Art. 100 provides clarification. In any particular installation, identification of that point can be made by the utility company and design personnel. The definition clarifies that the property line is not the determinant as to where NE Code rules must begin to be applied. This is particularly important in cases of multibuilding industrial complexes where the utility has distribution circuits on the property. See “Service Point” in Art. 100.

230.200 says that “service conductors and equipment used on circuits exceeding 600 V” must comply with all the rules in Art. 230 (including any “applicable provisions” that cover services up to 600 V). And Art. 100 says that for services up to 600 V, the “service conductors” are those conductors—whether on the primary or secondary of a step-down transformer or transformers—that carry current from the “service point” (where the utility connects to the customer’s wiring) to the service disconnecting means for a building or structure. See “Service Point” in Art. 100. All conductors between the defined points—“service point” and “service disconnecting means”—must comply with all requirements for service conductors, whether above or below 600 V.

Design and layout of any “service” are critically related to safety, adequacy, economics, and effective use of the whole system. It is absolutely essential that we know clearly and surely what circuits and equipment of any electrical system constitute the “service” and what parts of the system are not involved in the “service.” For instance, in a system with utility feed at 13.2 kV and step down to 480/277 V, the mandatory application of 230.95 requiring GFPE hinges on establishing whether the “service” is on the primary or secondary side of the transformers. If the secondary is the service, where the step-down transformer belongs to the utility and the “service point” is on its secondary, we have a mandatory need for GFPE and none of the Code rules on service would apply to any of the 13.2-kV circuits—regardless of their length or location. If the transformer belongs to the customer and the “service point” is on the primary side, the primary is the service, 230.95 does not require GFPE on services over 600 V phase-to-phase, all the primary circuit and equipment must comply with all of Art. 230, and the secondary circuits are feeders and do not have to comply with any of the service regulations. This potential loophole has been closed by the addition of comparable GFPE rules to feeder circuits in 215.10.

The whole problem involved here is complex and requires careful, individual study to see clearly the many interrelated considerations. Let us look at a few important things to note about Code definitions as given in Art. 100:

1. “Service conductors” run to the service disconnect of the premises supplied. Note that they run to “premises” and are not required to run to a “building.” The Code does not define the word premises, but a typical dictionary definition is “a tract of land, including its buildings” (Fig. 230-50).

2. “Service equipment” usually consists of “a circuit breaker or switch and fuses, and their accessories, located near the point of entrance of supply conductors to a building or other structure, or an otherwise defined area.” Note that the service equipment is the means of cutoff of the supply, and the service conductors may enter “a building” or “other structure” or a “defined area.” But, again, a service does not necessarily have to be to “a building.” It could be to such a “structure” as an outdoor switchgear or unit substation enclosure.

The wording in Art. 100 bases identification of “service conductors” as extending from the “service point.” Because of the definition of “service point,” it is essential to determine whether the transformers belong to the power company or the property owner.

Fig. 230-50. Where the transformer belongs to the utility, the “service point” is on its secondary and the secondary conductors are the service conductors to the building or structure. (Sec. 230.202 and Art. 100.)<None>

If a utility-owned transformer that handles the electrical load for a building is in a locked room or locked enclosure (accessible only to qualified persons) in the building and is fed, say, by an underground medium-voltage (over 600 V) utility line from outdoors, the secondary conductors from the transformer would be the “service conductors” to the building. And the switching and control devices (up to six CBs on fused switches) on the secondary would constitute the “service equipment” for the building. Under such a condition, if any of the secondary section “service disconnects” were rated 1000 A or more, at 480/277-V grounded wye, they would have to comply with 230.95, requiring GFP for the service disconnects.

However, if the utility made primary feed to a transformer or unit substation belonging to the owner, then the primary conductors would be the service conductors and the primary switch or CB would be the “service disconnect.” In that case, no GFPE would be needed on the “service disconnect” because 230.95 applies only up to 600 V, and there is no requirement for GFPE on medium-voltage services (Fig. 230-51). But in that case, there would be a need

Fig. 230-51. The primary is the “service” for any indoor transformer belonging to the owner and fed by utility line. (Sec. 230.200.)

for GFPE on the secondary section disconnects, even though they would not be “service disconnects”—and those are the same disconnects that might be subject to 230.95 if the transformer belonged to the utility. However, 215.10 or 240.13 may require such protection for these secondary section disconnects. (See also Fig. 230-52.)

230.202. Service-Entrance Conductors. This section specifies the minimum conductor size, that is, No. 6 in a raceway and No. 8 in a multiconductor cable. In addition, it indicates that only those wiring methods given in Secs. 300.37 (aboveground) and 300.50 (underground) may be used. That section gives the wiring methods that are acceptable for use as service-entrance conductors where it has been established that primary conductors (over 600 V) are the service conductors or where the secondary conductors are the service conductors and operate at more than 600 V. The basic conduits that may be used are rigid metal conduit, intermediate metal conduit, and rigid nonmetallic conduit. In

Fig. 230-52. The primary circuit must be taken as the “service conductors” where the “service point” is on the primary side of an outdoor transformer. Although GFPE is not required by 230.95 for the 480Y/277-V disconnects shown, it still is required because 215.10 applies instead.

addition, cable tray, cable bus, or “other identified” raceways or even type MC may be used. Note, too, that bare conductors, bare bus-work, or open runs of type MV are permitted as indicated. And the NEC no longer requires concrete encasement of the nonmetallic conduit.

Section 300.37 points out that cable tray systems are also acceptable for high-voltage services.” However, any such application for service work would still require application of the cable tray rules in 230.43 regarding segregation and marking if used for dual voltages. Medium-voltage (over 600 V) service-entrance cables may be used it they meet the requirements for the cables in Art. 426 and the rules for cable trays with medium voltage wiring in Art. 392. Details of this section are shown in Fig. 230-53.

230.204. Isolating Switches. An air-break isolating switch capable of visible verification of the blade position must be used ahead of an oil switch or an air, oil, vacuum, or sulfur hexafluoride CB used as a service disconnecting means, unless removable truck panels or metal-enclosed units are used providing disconnection of all live parts in the removed position. In addition, such removable equipment must not be openable unless the circuit is disconnected. This line-side disconnect ensures safety to personnel in maintenance (Fig. 230-54). Part (D) requires a grounding connection for an isolating switch, as in Fig. 230-55.

230.205. Disconnecting Means. In part (A), the basic rule requires a high-voltage service disconnecting means to be located “outside and within sight of, or inside nearest the point of entrance of, the service conductors” into the building or structure being supplied—as for 600-V equipment in 230.70. A new provision in 2008 allows this disconnecting means to be located where it is not readily accessible if part of an “overhead or underground primary distribution system.” The intent was to recognize the customary load break switches at the top of utility poles. Now it is true that the switch mechanism itself is not readily accessible. However, it is operable through a mechanical linkage at the pole base, and although mechanical, this meets all the provisions of 230.205(C) in that the switch is at a separate structure and operated remotely.

The real problem with this wording, however, is that it avoids compliance with the rest of 230.205(C), which is that any such remote disconnecting provision be located in a readily accessible location. In effect, however unconscionable, it may now be considered acceptable to install a pole-top switch with no linkage to the pole base, thereby relying on personnel working with a hot stick out of a bucket truck to open the switch. This was very unlikely to have been the intent.

Part (B), covering the electrical fault characteristics, requires that the service disconnect be capable of closing, safely and effectively, on a fault equal to or greater than the maximum short-circuit current that is available at the line terminals of the disconnect. The last sentence notes that where fuses are used within the disconnect or in conjunction with it, the fuse characteristics may contribute to fault-closing rating of the disconnect. This provision recognizes that some medium voltage fuses have current-limiting characteristics, and that having them in place will make it possible to close the switch safely. This might be seen as a modification of 110.9, which normally requires equipment that will interrupt circuits under fault conditions to be rated for the available fault current as their supply terminals. In this case, however, there is a distinction drawn between a fault

Fig. 230-53. Provisions for service conductors rated over 600 V (refer to subpart letter identification of rules). (Sec. 230.202.)

Fig. 230-54. Isolating switch may be needed to kill line terminals of service disconnect. (Sec. 230.204.)

Fig. 230-55. One method for grounding the load side of an open isolating switch. (Sec. 230.204.)

clearing rating, which the fuse will have and is normally quite high, and covered in 230.208, and a fault closing rating, to which the fuse may safely contribute.

230.208. Protection Requirements. Service conductors operating at voltages over 600 V must have a short-circuit (not overload) device in each ungrounded conductor, installed either (1) on load side of service disconnect, or (2) as an integral part of the service disconnect.

All devices must be able to detect and interrupt all values of current in excess of their rating or trip setting, which must be as shown in Fig. 230-56.

Fig. 230-56. Maximum permitted rating or setting of high-voltage overcurrent protection for service. (Sec. 230.208.)

The difference between 300 percent for fuses and 600 percent for CBs is explained as follows:

The American National Standards Institute (ANSI) publishes standards for power fuses. The continuous-current ratings of power fuses are given with the letter “E” following the number of continuous amps—for instance, 65E or 200E or 400E. The letter “E” indicates that the fuse has a melting time-current characteristic in accordance with the standard for E-rated fuses:

The melting time-current characteristics of fuse units, refill units, and links for power fuses shall be as follows:

(1) The current-responsive element with ratings 100 amperes or below shall melt in 300 seconds at an rms current within the range of 200 or 240 percent of the continuous current rating of the fuse unit, refill unit, or fuse link.

(2) The current-responsive element with ratings above 100 amperes shall melt in 600 seconds at an rms current within the range of 220 to 264 percent of the continuous current rating of the fuse unit, refill unit, or fuse link.

(3) The melting time-current characteristic of a power fuse at any current higher than the 200 to 240 or 264 percent specified in (1) or (2) above shall be shown by each manufacturer’s published time current curves, since the current-responsive element is a distinctive feature of each manufacturer.

(4) For any given melting time, the maximum steady-state rms current shall not exceed the minimum by more than 20 percent.

The fact that E-rated fuses are given melting times at 200 percent or more of their continuous-current rating explains why NE Code 230.208 and 240.100 set 300 percent of conductor ampacity as the maximum fuse rating but permit CBs up to 600 percent. In effect, the 300 percent for fuses times 2 (200 percent) becomes 600 percent—the same as for CBs.

Part (B) of this section permits overcurrent protection for services over 600 V to be loaded up to 100 percent of its rating even on continuous loads (operating for periods of 3 h or more). The greater spacings in medium-voltage equipment permit this latitude safely; see also 110.40 for allowances to use 90°C ratings on medium-voltage terminations for the same reason.

ARTICLE 240. OVERCURRENT PROTECTION

240.1. Scope. For any electrical system, required current-carrying capacities are determined for the various circuits—feeders, subfeeders, and branch circuits. Then these required capacities are converted into standard circuit conductors which have sufficient current-carrying capacities based on the size of the conductors, the type of insulation on the conductors, the ambient temperature at the place of installation, the number of conductors in each conduit, the type and continuity of load, and judicious determination of spare capacity to meet future load growth. Or if busway, armored cable, or other cable assemblies are to be used, similar considerations go into selection of conductors with required current-carrying capacities. In any case, the next step is to provide overcurrent protection for each and every circuit:

The overcurrent device for conductors or equipment must automatically open the circuit it protects if the current flowing in that circuit reaches a value which will cause an excessive or dangerous temperature in the conductor or conductor installation.

Overcurrent protection for conductors must also be rated for safe operation at the level of fault current obtainable at the point of their application. Every fuse and circuit breaker for short-circuit protection must be applied in such a way that the fault current produced by a bolted short circuit on its lead terminals will not damage or destroy the device. Specifically this requires that a short-circuit overcurrent device have a proven interrupting capacity at least equal to the current which the electrical system can deliver into a short on its line terminals. That is, the calculation for the short-circuit interrupting rating must not include the impedance of the device itself. That impedance may only be applied to the calculation for the next device downstream.

But safe application of a protective device does not stop with adequate interrupting capacity for its own use at the point of installation in the system. The speed of operation of the device must then be analyzed in relation to the thermal and magnetic energy which the device permits to flow in the faulted circuit. A very important consideration is the provision of conductor size to meet the potential heating load of short-circuit currents in cables. With expanded use of circuit-breaker overcurrent protection, coordination of protection from loads back to the source has introduced time delays in operation of overcurrent devices. Cables in such systems must be able to withstand any impressed short-circuit currents for the durations of overcurrent delay. For example, a motor circuit to a 100-hp motor might be required to carry as much as 15,000 A for a number of seconds. To limit damage to the cable due to heating effect, a much larger size conductor than necessary for the load current alone may be required.

A device may be able to break a given short-circuit current without damaging itself in the operation; but in the time it takes to open the faulted circuit, enough energy may get through to damage or destroy other equipment in series with the fault. This other equipment might be a cable or busway or a switch or motor controller—any circuit component which simply cannot withstand the few cycles of short-circuit current which flows in the period of time between initiation of the fault and interruption of the current flow.

The fine-print note (FPN) following Sec. 240.1 often raises questions about the approved use of conductors and overcurrent protection to withstand faults.

example Assume a panelboard with 20-A breakers rated 10,000 AIR (ampere interrupting rating) and No. 12 copper branch-circuit wiring. Available fault current at the point of breaker application is 8000 A. The short-circuit withstand capability of a No. 12 copper conductor with plastic or polyethylene insulation rated 60°C would be approximately 3000 A of fault or short-circuit current for one cycle.

question: Assuming that the CB (circuit breaker) will take at least one cycle to operate, would use of the conductor where exposed to 8000 A violate Sec. 110.9 or 110.10? These sections state that overcurrent protection for conductors and equipment is provided for the purpose of opening the electrical circuit if the current reaches a value which will cause an excessive or dangerous temperature in the conductor or conductor insulation. The 8000-A available fault current would seem to call for use of conductors with that rating of short-circuit withstand. This could mean that branch-circuit wiring from all 20-A CBs in this panelboard must be No. 6 copper (the next larger size suitable for an 8000-A fault current).

answer: As noted in UL Standard 489, a CB is required to operate safely in a circuit where the available fault current is up to the short-circuit current value for which the breaker is rated. The CB must clear the fault without damage to the insulation of conductors of proper size for the rating of the CB. A UL-listed, 20-A breaker is, therefore, tested and rated to be used with 20-A-rated wire (say, No. 12 THW) and will protect the wire in accordance with 240.4 when applied at a point in a circuit where the short-circuit current available does not exceed the value for which the breaker is rated. This is also true of a 15-A breaker on No. 14 (15-A) wire, for a 30-A breaker on No. 10 (30-A) wire, and all wire sizes.

UL 489 states:

A circuit breaker shall perform successfully when operated under conditions as described in paragraphs 21.2 and 21.3. There shall be no electrical or mechanical breakdown of the device, and the fuse that is indicated in paragraph 12.16 shall not have cleared. Cotton indicators as described in paragraphs 21.4 and 21.6 shall not be ignited. There shall be no damage to the insulation on conductors used to wire the device. After the final operation, the circuit breaker shall have continuity in the closed position at rated voltage.

240.2. Definitions. Here the Code provides a number of additional definitions that apply to this article on overcurrent protection. These definitions must be considered when interpreting the requirements given in this Code article. They will be referred to in the context of subsequent discussion of the relevant topics.

240.3. Other Articles. Here the Code reminds us that the rules given in Art. 240 are essentially general requirements for conductor protection. The individual articles for the equipment specifically indicated by this section have overcurrent protection requirements that are different from the “general rules” given in Art. 240 for protection of conductors and flexible cords. Where installing over-current protection for circuits for the equipment, and in the locations, identified here, the rules for overcurrent protection given in the indicated articles supersede the requirements given in Art. 240. In effect, compliance with those rules satisfies the rule of 240.3.

240.4. Protection of Conductors. Aside from flexible cords and fixture wires, conductors for all other circuits must conform to the rules of 240.4.

Clearly, the rule wants overcurrent devices to prevent conductors from being subjected to currents in excess of the ampacity values for which the conductors are rated by 310.15 and Tables 310.16 through 310.21.

The wording mentions 310.15, which includes Tables 310.16 through 310.21. That is important because it points out that when conductors have their ampacities derated because of conduit fill [310.15(B)(2)] or because of elevated ambient temperature, the conductors must be protected at the derated ampacities and not at the values given in the tables.

The basic rule of Sec. 240.4(A) represents a basic concept in Code application. When conductors supply a load to which loss of power would create a hazard, this rule states it is not necessary to provide “overload protection” for such conductors, but “short-circuit protection” must be provided. By “overload protection,” this means “protection at the conductors’ ampacity”—that is, protection that would prevent overload (Fig. 240-1).

Fig. 240-1. If “overload protection” creates a hazard, it may be eliminated. (Sec. 240.4.)

Several points should be noted about this rule.

1. This requirement is reserved only for applications where circuit opening on “overload” would be more objectionable than the overload itself, “such as in a material handling magnet circuit.” In that example mentioned in the rule, loss of power to such a magnet while it is lifting a heavy load of steel would cause the steel to fall and would certainly be a serious hazard to personnel working below or near the lifting magnet. To minimize the hazard created by such power loss, the circuit to it need not be protected at the conductor ampacity. A higher value of protection may be used—letting the circuit sustain an overload rather than opening on it and dropping the steel. Because such lifting operations are usually short-time, intermittent tasks, occasional overload is far less a safety concern than the dropping of the magnet’s load.

2. The rule to eliminate only “overload protection” is not limited to a lifting-magnet circuit, which is mentioned simply as an example. Other electrical applications that present a similar concern for “hazard” would be equally open to use of this rule. Fire pump circuits are required to implement this principle.

3. Although 240.4(A) allows elimination of overload protection and requires short-circuit protection, it gives no guidance on selecting the actual rating of protection that must be used. For such circuits, fuses or a CB rated, say, 200 to 400 percent of the full-load operating current would give freedom from overload opening. Of course, the protective device ought to be selected with as low a rating as would be compatible with the operating characteristics of the electrical load. And it must have sufficient interrupting capacity for the circuit’s available short-circuit current.

4. Finally, it should be noted that this is not a mandatory rule but a permissible application. It says “. . . overload protection shall not be required . . .”; it does not say that overload protection “shall not be used.” Overload protection may be used, or it may be eliminated. Obviously, careful study should always go into application of this requirement.

Specifically, the general rule is that the device must be rated to protect conductors in accordance with their safe allowable current-carrying capacities. Of course, there will be cases where standard ampere ratings and settings of over-current devices will not correspond with conductor capacities. In such cases, part (B) permits the next larger standard size of overcurrent device to be used where the rating of the protective device is 800 A or less, unless the circuit in question is a multioutlet receptacle circuit for cord- and plug-connected portable loads, in which case the next smaller standard size overcurrent device must be used. Therefore, a basic guide to effective selection of the amp rating of overcurrent devices for any feeder or service application is given in various subsections [(A) through (G)].

For example, if a circuit conductor of, say, 500-kcmil THW copper (not more than three in a conduit at not over 86°F [30°C] ambient) satisfies design requirements and NE Code rules for a particular load current not in excess of the conductor’s table ampacity of 380 A, then the conductor may be protected by a 400-A rated fuse or CB.

240.6, which gives the “Standard Ampere Ratings” of protective devices to correspond to the word “standard” in part (B), shows devices rated at 350 and 400 A, but none at 380 A. In such a case, the NE Code accepts a 400-A-rated device as “the next higher standard device rating” above the conductor ampacity of 380 A.

But, such a 400-A device would permit load increase above the 380 A that is the safe maximum limit for the conductor. Better conductor protection could be achieved by using a 350-A-rated device, which will prevent such overload.

For application of fuses and CBs, parts (B) and (C) have this effect:

1. If the ampacity of a conductor does not correspond to the rating of a standard-size fuse, the next larger rating of fuse may be used only where that rating is 800 A or less. Over 800 A, the next smaller fuse must be used, as covered in part (C). For any circuit over 800 A, 240.4(C) prohibits the use of “the next higher standard” rating of protective device (fuse or CB) when the ampacity of the circuit conductors does not correspond with a standard ampere rating of fuse or CB. The rating of the protection may not exceed the conductor ampacity. Although it would be acceptable to use a protective device of the next lower standard rating (from 240.6) below the conductor ampacity, there are many times when greater use of the conductor ampacity may be made by using a fuse or CB of rating lower than the conductor ampacity but not as low as the next lower standard rating. Listed fuses and CBs are made with ratings between the standard values shown in 240.6.

For example, if the ampacity of conductors for a feeder circuit is calculated to be 1540 A, 240.4(C) does not permit protecting such a conductor by using the next higher standard rating above 1540 to 1600 A. The next lower standard rating of fuse or CB shown in 240.6 is 1200 A. Such protection could be used, but that would sacrifice 340 A (1540 minus 1200) of conductor ampacity. Because listed 1500-A protective devices are available and would provide for effective use of almost all the conductor’s 1540-A capacity, this rule specifically recognizes such an application as safe and sound practice. Such application is specifically recognized by the second sentence of 240.6.

In general, 240.6 is not intended to require that all fuses or CBs be of the standard ratings shown. Intermediate values of protective device ratings may be used, provided all Code rules on protection—especially the basic first sentence of 240.4, which requires conductors to be protected at their ampacities—are satisfied (Fig. 240-2).

2. A nonadjustable-trip breaker (one without overload trip adjustment above its rating—although it may have adjustable short-circuit trip) must be rated in accordance with the current-carrying capacity of the conductors it protects—except that the next higher standard rating of CB may be used if the ampacity of the conductor does not correspond to a standard unit rating. In such a case, the next higher standard setting may be used only where the rating is 800 A or less. An example of such application is shown in Fig. 240-1, where a nonadjustable CB with a rating of 1200 A is used to protect the conductors of a feeder circuit which are rated at 1140 A. As shown there, use of that size CB to protect a circuit rated at 1140 A (3 × 380 A = 1140 A) clearly violates 240.4(C) because the CB is the next higher rating above the ampacity of the conductors—on a circuit rated over 800 A. With a feeder circuit as shown (three 500-kcmil THW, each rated at 380 A), the CB must not be rated over 1140 A. A standard 1000-A CB would satisfy the Code rule—being the next lower rated protective device from 240.6. Or a 1100-A fuse could be used. Of course, if 500-kcmil THHN or XHHW conductors are used instead of THW conductors, then each 500 is rated at 430 A, three per phase would give the circuit an ampacity of 1290 A (3 × 430), and the 1200-A CB would satisfy the basic rule in 240.4(C). But, given that 500-kcmil conductors would be operating at 90°C when carrying the 430 A of current, such conductors could never be loaded to that value as there is no equipment rated for use with conductors operating at 90°C. To satisfy the termination temperature limitations of 110.14(C), the load would be prohibited from exceeding the 75°C value, or 3 × 380 A = 1140 A. Alternatively, a circuit breaker listed for terminations operating at 90°C could be used. These are only available in very large frame sizes, such as those in this example.

Fig. 240-2. Protection in accordance with 240.4(C) may use standard or nonstandard rated fuses or circuit breakers. And, smaller conductors are considered protected as covered in 240.4(D). (Secs. 240.4 and 240.6.)

It should be noted, however, that 240.4(B) requires that the rating of overcurrent protection must never exceed the ampacity of circuit conductors supplying one or more receptacle outlets on a branch circuit with more than one outlet. This wording in 240.4(B) coordinates with the rules described under 210.19(A)(2) on conductor ampacity. The effect of that rule is to require that the rating of the overcurrent protection must not exceed the Code-table ampacity (NEC Table 310.16) or the derated ampacity dictated by 310.15(B)(2) for any conductor of a multioutlet branch circuit supplying any receptacles for cord-and plug-connected portable loads. If a standard rating of fuse or CB does not match the ampacity (or derated ampacity) of such a circuit, the next lower standard rating of protective device must be used. But, where branch-circuit conductors of an individual circuit to a single load or a multioutlet circuit supply only fixed connected (hard-wired) loads—such as lighting outlets or permanently connected appliances—the next larger standard rating of protective device may be used in those cases where the ampacity (or derated ampacity) of the conductor does not correspond to a standard rating of protective device—but, again, that is permitted only up to 800 A, above which the next lower rating of fuse or CB must be used, as described under 210.19(A)(2).

The rules of 240.4 must also be correlated with the requirement for minimum conductor size where continuous loading or a combination of continuous and noncontinuous loading is supplied. Where such loads are supplied, 210.19(A), 210.20(A), 215.2(A), 215.3(A), and 230.42 require that additional capacity be provided where the branch-circuit, feeder, or service conductors and overcurrent devices supply continuous loads. After that minimum size has been established, the overcurrent protective device must be rated such that it either protects the conductors in accordance with their ampacity, or is the next larger rated over-current device—up to 800 A. Above 800 A, 240.4(C) would mandate use of the next smaller rated overcurrent device, which may not be adequately rated to supply the continuous load. Careful correlation of the rules here and in 210.19, 210.20, 215.2, 215.3, and 230.42 is especially important to ensure the selected conductors and OC protection are properly rated.

At the end of this book, as part of the detailed coverage of ampacity calculations, all of these code requirements are integrated in one location. The coverage focuses on Annex D, Example D3(a), which is the new example devoted to ampacity calculations as distinguished from load calculations.

Part (D) of 240.4 covers a long-standing requirement for protection of the smaller sizes of conductors, that is, No. 14, No. 12, and No. 10. Although such conductors have greater ampacities, as shown in Table 310.16, this rule requires that the maximum rating of overcurrent protection be 15, 20, and 30 A, respectively. This limitation on the rating of overcurrent devices is related to the fact that listed overcurrent devices cannot protect against conductor damage under short-circuit testing where, say, a No. 12 copper THW conductor is protected by a 25-A CB. Although the No. 12 has an ampacity of 25 A and can carry that current when used under the conditions described in the heading of Table 310.16, a 25-A CB will not operate fast enough to prevent the conductor from burning open during the short-circuit test. It was established that lower-rated breakers, such as a 20-A CB, can protect the No. 12 conductor from damage under short-circuit conditions. Because this is not a problem with CBs rated for protection of conductor sizes No. 8 and larger, only No. 14, No. 12, and No. 10, copper, as well as No. 12 and No. 10 aluminum and copper-clad aluminum conductors, are specially limited with regard to the maximum rating of their overcurrent protection, where they are required to be protected in accordance with their ampacity.

In the 2008 NEC, this part of 240.4 has been expanded to cover even smaller conductors (18 and 16 AWG) and the allowable overcurrent devices that can be used to protect them when they are not considered to be protected by ordinary branch-circuit protective devices due to special provisions in various Code rules. A good example would be a field-assembled extension cord set, as covered in 240.5(B)(4). If the cord is made from 16 AWG cord, it can be used on conventional 15- and 20-A branch circuits even though it is smaller than 14 AWG.

240.4(G) gives a list of “specific conductor applications” that are exempt from the basic rules of 240.4. For example, this rule refers the matter of protecting motor-control circuits to Art. 430 on motors.

Table 240.4(G) also applies to the protection of the remote-control circuit that energizes the operating coil of a magnetic contactor, as distinguished from a magnetic motor starter (Fig. 240-3).

725.45(C) covers control wires for magnetic contactors used for control of lighting or heating loads, but not motor loads. 430.72 covers that requirement for motor-control circuits. In Fig. 240-4, the remote-control conductors may be considered properly protected by the branch-circuit overcurrent devices (A) if these devices are rated or set at not more than 300 percent of (3 times) the