Sizing a neutral conductor starts with calculating the maximum unbalanced load: the highest current that could flow between the neutral and any single phase conductor. NEC Section 220.61 governs this calculation for feeders and services, and it applies whether you used the standard or optional method for your overall load calculation. The neutral carries only the imbalance between phases, so in a perfectly balanced system it would carry zero current. Real-world systems are never perfectly balanced, and certain types of loads can push neutral current even higher than phase current.
The Basic Rule: Maximum Unbalanced Load
NEC 220.61(A) defines the neutral load as the maximum net calculated load between the neutral conductor and any one ungrounded (hot) conductor. For a single-phase, three-wire system (like a typical 120/240V residential panel), this means you look at which leg has the heavier load. The difference between the two legs is what flows on the neutral, but you size the neutral for the full load on the heavier leg, because that represents the worst-case scenario if one side were fully loaded and the other carried nothing.
For three-phase, four-wire systems, the math gets more involved. The neutral current in an unbalanced three-phase system can be calculated with this formula:
In = √(A² + B² + C² − AB − BC − AC)
Where A, B, and C are the currents on each phase. You construct this from the vector sum of all three phases: since the phase currents are offset by 120 degrees, the neutral current equals the negative of the vector sum of all three phase currents. In practice, many designers calculate the worst-case unbalance by assuming the heaviest loaded phase carries its full calculated load while the others carry less.
The 70% Demand Factor Reduction
NEC 220.61(B)(2) allows a useful reduction for larger systems. On feeders or services supplied from a three-wire single-phase, four-wire three-phase, or similar multi-wire system, you can apply a 70% demand factor to the portion of the unbalanced neutral load that exceeds 200 amps. So the first 200 amps of neutral load are calculated at 100%, and everything above that is multiplied by 0.70.
For example, if your calculated maximum unbalanced load is 350 amps, the neutral sizing calculation would be: 200 + (150 × 0.70) = 305 amps. This reduction reflects the statistical reality that all circuits rarely hit peak load simultaneously, keeping the neutral from being unnecessarily oversized on large services.
There is one major exception to this reduction, and it catches people off guard.
Non-Linear Loads and Harmonics
NEC 220.61(C) prohibits applying the 70% demand factor reduction to any portion of the neutral load consisting of non-linear loads on a four-wire, wye-connected, three-phase system. This is one of the most critical rules in neutral sizing, and it exists because of harmonics.
Non-linear loads, such as computers, LED drivers, variable-frequency drives, and switch-mode power supplies, draw current in pulses rather than smooth sine waves. Those pulses create harmonic currents at frequencies that are multiples of the standard 60 Hz. The most problematic are “triplen” harmonics (3rd, 9th, 15th). In a four-wire wye system, triplen harmonic currents from all three phases are in-phase with each other. Instead of canceling out on the neutral the way balanced fundamental currents do, they add together. This means the neutral can carry significantly more current than any individual phase conductor.
Sizing the neutral for these systems requires several steps:
- Inventory your loads. Determine what portion of the feeder load is non-linear. The NEC doesn’t specify an exact percentage threshold, so this involves engineering judgment and coordination with the local authority having jurisdiction.
- Measure actual conditions when possible. A true RMS ammeter or power-quality analyzer will capture the real heating effect of harmonic-laden current. Standard averaging meters underread harmonic current and give dangerously optimistic numbers.
- Keep the neutral at full size (or larger). Do not reduce the neutral below the phase conductors when serving heavy non-linear loads. Many engineers upsize the neutral to 200% of the phase conductor ampacity in extreme cases, such as data centers or large office buildings full of electronic equipment.
- Count the neutral as a current-carrying conductor. When harmonic currents are present, the NEC requires the neutral to be counted for ampacity adjustment purposes. In a typical three-phase circuit, this bumps you from three current-carrying conductors to four, which triggers a derating factor. Four to six current-carrying conductors in a raceway typically require an 80% ampacity adjustment.
Voltage Drop on Long Runs
Code-compliant ampacity is only half the equation. On long conductor runs, voltage drop can force you to upsize the neutral beyond what the load calculation alone requires. The NEC recommends (and many jurisdictions enforce) a maximum 3% voltage drop on service conductors, 2% on feeders or branch circuits, and no more than 5% combined from the service entrance to the final outlet.
Voltage drop depends on conductor length, material (copper vs. aluminum), and the current flowing through it. Because the neutral carries return current, it contributes to the total circuit voltage drop just like the phase conductors do. If your phase conductors needed to be upsized for voltage drop, the neutral generally needs the same treatment. Calculate voltage drop for the full circuit loop, including the neutral return path, and size accordingly.
Parallel Neutral Conductors
For high-ampacity installations, NEC 310.10(H) permits running conductors in parallel for each phase and for the neutral. The minimum size for parallel conductors is 1/0 AWG. Each parallel set must use conductors of the same length, material, size, and insulation type, and they must be terminated in the same manner. This ensures current divides equally among the parallel paths. If one conductor in a parallel set is longer or a different size, it will carry a different share of the current, creating an imbalance that could overheat one conductor while underloading another.
When calculating the neutral size in a parallel installation, determine the total required neutral ampacity first, then divide by the number of parallel conductors to find the per-conductor ampacity requirement. Each individual conductor must meet the minimum 1/0 AWG size regardless of how low the per-conductor current works out.
Identifying the Neutral Conductor
Once you’ve sized the neutral, proper identification matters for safety and inspection. For conductors 4 AWG and larger, you can use a conductor with a continuous white or gray outer finish from the manufacturer, or you can mark it at each termination point with white or gray tape that encircles the full circumference of the insulation. Most electricians use electrical tape for this. Smaller conductors (6 AWG and below) typically come with white or gray insulation from the factory.
Putting It All Together
A practical neutral sizing workflow looks like this: first, calculate the maximum unbalanced load per 220.61(A) using your load schedule. Second, apply the 70% demand factor to any portion exceeding 200 amps, but only for the linear portion of the load. Third, assess non-linear loads and keep that portion of the neutral at full calculated value with no reduction. Fourth, check whether harmonics require counting the neutral as a current-carrying conductor and apply the appropriate ampacity adjustment. Fifth, run a voltage drop calculation for the full circuit length and upsize if needed.
In a straightforward residential panel with mostly resistive and motor loads, the neutral often ends up smaller than the phase conductors because the 70% reduction applies and the loads are reasonably balanced. In a commercial building loaded with electronic equipment, the neutral may need to be the same size as the phase conductors or larger. The type of load, not just the amount, drives the final answer.