Voltage drop is the least dramatic problem in power distribution. Nothing arcs, nothing trips on the first day, and the symptoms masquerade as equipment problems for years. It is also one of the cheapest problems to prevent on paper and one of the most expensive to correct in conduit — which is exactly the trade a voltage drop analysis exists to catch early.
What is voltage drop?
Every conductor has impedance. Current through impedance costs voltage, so the voltage that arrives at the load is the source voltage minus everything spent along the way: the service conductors, the transformer, the feeder, the branch circuit. Each piece is small. The load sees the sum.
The consequences land on the equipment. An induction motor at reduced terminal voltage draws more current to deliver the same torque, so it runs hotter and its insulation ages faster — and its starting torque falls off roughly with the square of the voltage, which is why marginal voltage shows up first as hard starting. Drives ride through some sag and then trip on undervoltage. Contactors drop out. None of it presents as "a voltage problem"; it presents as flaky equipment.
Where do the 3% and 5% figures come from?
From informational notes in the NEC — roughly 3% on a branch circuit and 5% total for feeder plus branch. Informational notes are guidance, not enforceable requirements, which surprises people who have treated 5% as law for their whole careers. A few specific articles do impose firm voltage requirements for particular systems — fire pump circuits are the classic example — and energy codes in some jurisdictions add enforceable conductor-sizing rules of their own.
The engineering reality: the guidance numbers are a sensible envelope, and the true limit is whatever your equipment tolerates. A drive with a tight undervoltage window, a motor at the end of a long feeder run, or a control circuit fed from the same panel as a large cyclic load will all complain before the 5% total is reached on paper.
How is voltage drop calculated?
For one conductor run, the hand method is standard: current times conductor impedance for the length, doubled for the out-and-back path, with a √3 factor instead of 2 for balanced three-phase circuits. Conductor resistance values come straight from the NEC's conductor tables, and for larger conductors the reactance term matters as much as resistance — a detail hand calculations tend to drop.
That answers a single run in isolation. The question a facility actually needs answered is cumulative: the load at the end of a branch circuit also sits behind the feeder's drop and the transformer's regulation, all at once, and under the loading that actually occurs rather than the connected total on the panel schedule. Stacking those correctly across every path in a building is a load flow study — the same model computes loading on every transformer and feeder against its rating, which is the other half of the capacity question. (How that analysis works and when to run it is its own article: Load Flow Analysis: Sizing Your Electrical System for Growth.)
When do rules of thumb fail?
Predictably, in the same handful of situations:
- Long runs to real loads. Wells, remote buildings, parking-lot circuits, crane rails — anywhere length dominates, the branch-circuit habit of "upsize one gauge" stops being enough.
- Stacked drops. The feeder was sized to 2%, the branch to 3%, the transformer sags under load — each number defensible alone, and the load at the end sees the sum.
- Motor starting. Running-load calculations say nothing about the voltage dip during a large motor's start, which is when drives trip and contactors chatter. Starting studies are a load flow question, not a conductor table lookup.
- Growth by accretion. Every added load raises the current in conductors sized for a smaller building. The circuits that were fine at commissioning drift toward marginal with each addition, with no single change to blame.
What are the fixes?
Once the analysis shows where the voltage goes, the options are ordinary and mostly cheap in design, expensive in retrofit: larger conductors on the runs that matter, transformer tap adjustments to center the delivered voltage, splitting long circuits, moving large loads closer to the source electrically, or distributing at a higher voltage and transforming down near the load. Which one is right is a numbers question — the analysis prices them against each other.
If your facility shows the symptoms in the FAQ above, or you are about to add significant load to a system nobody has modeled, the load flow and voltage drop service page covers what the study includes. Sizing conductors correctly on paper costs a few hours of engineering; discovering the feeder is undersized after the concrete is poured costs considerably more.