The neutral question gets answered before any table is opened, and the rest of the item rides on it

How the exam asks about the count

Sixteen of the 68 conductor sizing questions in our bank turn on a count of current-carrying conductors, and 22 do across the full 500 once the conduit items are added. Five of the 68 name a neutral outright, and six questions across the bank hand you a nonlinear or harmonic load and wait to see what you do with the neutral.

That is a modest count carrying an outsized consequence, because the count is upstream of everything else. A misread neutral produces a clean wrong number that no later step contradicts: the adjustment lookup accepts it, the multiplication accepts it, the termination comparison accepts it, and the answer comes out wrong with every subsequent step done correctly.

It costs time as well. Work the count last, after the lookups, and a wrong count means doing the lookups again. Work it first and it costs fifteen seconds once. The journeyman calculations portion gives you 110 minutes for 26 items, about 4.2 minutes each, and a repeated lookup is most of one item's budget spent twice.

What the 2026 edition moved

310.15(E) was rebuilt for the 2026 edition into a two-way switch. 310.15(E)(1) is a short positive list of the circuits whose neutral is current-carrying. 310.15(E)(2) picks up everything the list left behind, where a neutral carrying only the unbalanced current from the other conductors of the same circuit stays out of the count. Find which side of the switch the stem put you on and the answer follows.

Read the qualifier ahead of the exclusion, because the exclusion reaches only circuits the positive list declined to take. It can never be used to argue a neutral back out of the count on a two-wire circuit or on three wires off a wye. Candidates who learned this from an older book tend to carry the exclusion first and the list second, and that ordering is what produces the wrong answer.

The timing matters in Texas. 16 TAC 73.100 adopts the 2026 National Electrical Code effective 1 September 2026, and PSI referenced the examinations to the same edition from that day. So a candidate sitting in the autumn of 2026 is answering against the restructured section while most study material still describes the older shape. The edition question is worth settling against your own exam date.

The question that settles it

Ask what kind of circuit it is and what kind of load is on it. 310.15(E) is written entirely in those terms and it names no measured value anywhere, which is exactly why the rule works at the design stage, before anything has been energized and before anybody could take a reading.

So a neutral carrying 30 amperes can sit outside the count and a neutral carrying almost nothing can sit inside it. The word balanced is where the confusion starts, because balance is a property of the loads at a moment, while the section asks whether the neutral structurally carries current the ungrounded conductors are already accounting for.

The physics underneath is one sentence, and it lets you rebuild every case from scratch. The adjustment factors exist because conductors bundled together struggle to shed their heat, so a conductor adding no heat adds nothing to the count. A neutral carrying only the unbalanced current of its own circuit is carrying current one of the ungrounded conductors is therefore carrying less of, the total heat in the raceway holds where the table assumed it would, and there is nothing to add.

The three circuits on the list

All three sit in 310.15(E)(1), and the code states them as circuits. That is why a well-built stem always names the system, and why a stem that names the system is telling you the answer if you are listening for it.

A two-wire circuit

One ungrounded conductor and one neutral. The neutral carries exactly what the ungrounded conductor carries for as long as the circuit is energized, so both conductors count. This case is missed only by people applying the exclusion as a slogan.

Three wires off a four-wire wye

Two ungrounded conductors and the neutral, where the two phases sit 120 degrees apart. Add two equal currents 120 degrees apart and the result comes out equal in size to either one of them, so the neutral of that circuit carries full load current when the two phases are perfectly matched. Balance leaves this neutral busy.

A four-wire wye with a major portion of nonlinear load

All four counted, and this one is about harmonics. A nonlinear load draws current in pulses, those pulses carry a third harmonic component, and third harmonic currents from the three phases arrive at the neutral in step with one another and add up. Note which system the code names, because that is the whole reach of the rule: the nonlinear trigger attaches to a four-wire, three-phase wye circuit. Major portion is the code's own phrase and it is a judgment, so an exam stem will say it outright.

The two three-wire circuits that look alike

Hold the two pictures next to each other, because in a raceway they are indistinguishable and they count differently. Two legs 180 degrees apart and the neutral carries the difference between the two load currents. Two phases 120 degrees apart and the neutral carries about as much as either one. Same three wires, same pipe, opposite answers, and the only thing separating them is the clause in the stem naming the system.

A single-phase three-wire circuit feeding a floor of switching power supplies falls outside the nonlinear trigger, and the physics agrees with the drafting: two legs 180 degrees apart put their third harmonics 180 degrees apart as well, so those harmonics subtract in the neutral. What makes a load nonlinear is a switching supply in front of it, so electronic ballasts, LED drivers, computer power supplies and variable frequency drives all qualify, while a resistance heater and an incandescent lamp stay linear.

Which wrong answer people pick

Our bank records the reasoning behind each wrong option, and four of them repeat across this topic.

• The neutral counts in every case, because a real load is never perfectly balanced. The residual unbalance is exactly what the exclusion contemplates, and the nonlinear case is the special one that carrying this as a general rule gets you to.
• The neutral counts as half a conductor, since it carries the unbalance. The unbalance reasoning is sound and the fraction has nowhere to live, because a conductor is either counted or left out.
• The neutral counts above 100 amperes. The trigger is the kind of circuit and the kind of load, and the size of the load leaves it alone.
• The equipment grounding conductor joined the count. 310.15(F) keeps grounding and bonding conductors out of the adjustment entirely, and for a different reason from the neutral exclusion: a qualifying neutral is left out because its current is accounted for elsewhere in the same circuit, while a grounding conductor is left out because normal service puts no current on it at all.

Two further counting rules sit in the note printed under the adjustment table. Conductors connected to components that cannot be energized at the same time drop out of the count, and spare conductors stay in it, because a spare is one termination away from being a live circuit and the raceway will not know the difference.

Two worked counts, and the other count

Take the circuits one at a time. A raceway holds one three-phase, four-wire wye circuit supplying electronic power supplies and lighting ballasts, one two-wire circuit and one equipment grounding conductor. The wye circuit supplies nonlinear load, so its neutral counts with the three ungrounded conductors, which is four. The two-wire circuit contributes two. The grounding conductor stays out. The count is six.

Now change the mix. Two 120 volt two-wire circuits and one 120/240 volt three-wire circuit share a raceway with a single equipment grounding conductor, and every load is linear. The two-wire circuits contribute four, since in a two-wire circuit the grounded conductor carries the full circuit current. The three-wire circuit from a 120/240 volt system contributes two, because its neutral carries only the difference between the two legs. The count is six again, by a completely different route.

Keep this count apart from conduit fill, which counts everything in the raceway including the conductors excluded here. A bank item makes the difference bite: a raceway filled to 0.404 square inches against a permitted 0.406 takes a bare grounding conductor of 0.0134 square inches, and the fill fails at 0.4174 while the adjustment count stays where it was. One drawing, two counts, and a well-built item asks for both. The raceway area carries the fill side.

What to do before you sit

The count reappears one level up as well. In a feeder neutral calculation, a further 70 percent demand factor is available on the portion of the unbalanced load above 200 amperes, and 120.61 withholds it in two situations: a three-wire circuit of two ungrounded conductors and the neutral taken from a four-wire wye, and the portion of the load that is nonlinear and supplied from a four-wire wye. Both are cases where the neutral declines to behave the way the reduction assumes, so the same two facts you learned for the adjustment count answer a load calculation item as well.

• Name the system before you count anything. Two-wire, single-phase three-wire, four-wire wye, or three wires taken off one of those, and the answer follows from the name.
• Work the count first on every item, then open the table. Fifteen seconds spent there saves a repeated lookup on a portion that gives you about four minutes an item.
• Tab 310.15(E), 310.15(F) and the adjustment table together, along with Chapter 9 Table 1 for the fill count off the same drawing. Publisher tabs only, since homemade ones are turned away at the door.
• Drill the two lookalike three-wire circuits back to back until the system name does the work, because the derating arithmetic that follows is worthless on a count you got wrong.

The worked counts, with the section text tracked line by line and the adjustment applied afterward, sit in the question set.

Questions people ask

Is a neutral a current-carrying conductor?

It depends on the circuit, and a measurement settles none of it. NEC 310.15(E)(1) names three circuits whose neutral is considered current-carrying, and NEC 310.15(E)(2) says that for any circuit outside that list, a neutral carrying only the unbalanced current from the other conductors of the same circuit stays out of the count. So a neutral carrying 30 amperes can sit outside the count and a neutral carrying almost nothing can sit inside it. That is why the rule works at the design stage, before anything has been energized.

Does the neutral count on a balanced three-phase wye feeder?

With linear line-to-neutral loads it stays out, and the count is three. Three ungrounded conductors and a neutral share the raceway, the neutral carries only the imbalance, and the ampacity table already assumed that heat. This is the most common item on the subject in any bank, and the option offering four is there for the candidate who counted the neutral because it was physically in the pipe. Where a major portion of the load is nonlinear the answer flips to four, because harmonic current in the neutral declines to cancel.

When does the neutral count because of harmonics?

On a four-wire, three-phase wye circuit where the major portion of the load consists of nonlinear loads, which is the third entry in NEC 310.15(E)(1). Third harmonic currents from the three phases arrive at the neutral in step with one another and add up, so the neutral can end up carrying more than any single phase conductor feeding it. Note which system the code names, because a single-phase three-wire circuit feeding a floor of switching power supplies falls outside the trigger, and there the third harmonics sit 180 degrees apart and subtract.

Does the neutral count on a three-wire circuit off a wye?

Yes. NEC 310.15(E)(1) names a three-wire circuit of two ungrounded conductors and the neutral conductor of a four-wire, three-phase wye connected system, and the informational note under it says that neutral carries approximately the same current as the line-to-neutral load current of the other conductors. Two currents 120 degrees apart leave a sum equal in size to either one, so balance leaves this neutral busy. In a raceway it looks identical to a single-phase three-wire circuit and it counts differently, which is why a well-built stem always names the system.

Are grounding conductors counted for the adjustment?

NEC 310.15(F) keeps grounding and bonding conductors out of the count entirely, because normal service puts no current on them and the ampacity tables were built around a sustained thermal condition. Keep that reason separate from the neutral exclusion, since questions test the difference: a qualifying neutral is left out because its current is already accounted for elsewhere in the same circuit. Grounding conductors do still occupy space, so they go into the conduit fill calculation, which counts everything in the raceway and excludes nothing.

Do spare conductors go into the count?

Yes. The note printed under the adjustment table keeps spare conductors in the count, because a spare is one termination away from being a live circuit and the raceway will not know the difference. The same note takes conductors out where they are connected to components that cannot be energized at the same time. Those two rules sit with the table, outside 310.15(E), so a candidate who reads only the neutral section will meet them for the first time in an option list.