Voltage drop field guide for electrical crews

Voltage drop, and why the far end is what matters

Voltage drop is the voltage lost in a conductor while current flows through it. The supply might read 208 V at the panel, but the motor 300 ft away sees less, because the wire itself has resistance and that resistance turns part of your voltage into heat along the run. The number that matters is never the voltage at the source. It is the voltage at the load.

Four things drive it: how much current you push, how far you push it, the resistance of the conductor, and temperature, which raises that resistance as the wire warms. Get any of those wrong in the calculation and the answer is wrong in a way that looks fine on paper.

On the job the calculation is rarely the hard part. The trouble is using the plan distance instead of the routed length, the design current instead of the real continuous load, or a resistance value that does not match the conductor you actually pulled. Then nobody writes down which assumption drove the conductor size, so when the equipment at the end runs low six months later, there is no record to check.

What low voltage actually does to the equipment

Undervoltage does not announce itself. It shows up as symptoms that get blamed on the equipment first and the wiring last.

Motors are the classic case. A motor starved for voltage draws more current to make the same torque, runs hotter, and trips on overload or cooks the windings over time. Contactors chatter and weld their contacts. On a long feeder to a rooftop unit, the compressor that nuisance-trips on a hot afternoon is often a voltage-drop problem wearing a mechanical disguise.

Lighting dims and shifts color. Electronics and drives have a tighter tolerance than people expect, and many list a minimum operating voltage in the manual that is stricter than the 3 percent rule of thumb. PoE devices at the end of a long Cat6 run can drop below their power class and reset or fail to boot. The point is the same across all of it: the calculation protects the equipment, not the inspection.

How do you calculate voltage drop?

Use current, one-way length, and the conductor's resistance per 1000 ft. For a single-phase circuit, multiply current by two (the conductor and the return), by the one-way length, by the resistance per 1000 ft, then divide by 1000. For a three-phase circuit, replace the 2 with 1.732, the square root of three, because the phase relationship changes how the drops add.

The factor catches people. The 2 in the single-phase formula already accounts for the round trip, so you feed it the one-way length, not the loop. The 1.732 in the three-phase formula is not a round-trip factor at all. Both formulas take the one-way run.

Resistance comes from the conductor tables, not from memory. NEC Chapter 9, Table 8 gives DC resistance per 1000 ft, which is close enough for most branch and feeder work. Table 9 gives AC resistance and reactance for conductors in raceway, which matters more on large conductors and long runs where reactance is no longer negligible. For most field checks, Table 8 is the practical source.

I

Load current in amps, the continuous design current the conductor actually carries

L

One-way conductor length in feet, measured along the routed path

R

Conductor resistance in ohms per 1000 ft, from NEC Chapter 9, Table 8 or Table 9

Field example: 208 V feeder over a 150 ft run

A three-phase 208 V feeder carries 60 A over a 150 ft one-way route in copper #4 AWG. From NEC Chapter 9, Table 8, the DC resistance of stranded #4 copper is 0.308 ohms per 1000 ft.

Run the three-phase formula: 1.732 times 60 A times 150 ft times 0.308, divided by 1000, gives 4.80 V dropped. As a percentage of 208 V, that is 2.31 percent, inside a 3 percent branch target but with less margin than the round number suggests. Add the feeder drop ahead of it and you can be past 3 percent before the branch circuit even starts.

Now change one input and watch the answer move. Stretch the route to 220 ft and the drop climbs to about 3.4 percent on its own. Swap the copper for aluminum #4 at roughly 0.49 ohms per 1000 ft and the same 150 ft run jumps to about 3.7 percent. This is why the conductor decision lives in the calculation, not in habit.

What voltage drop is acceptable?

Acceptable voltage drop is commonly held to 3 percent on a branch circuit and 5 percent total across feeder plus branch. Those numbers come from informational notes in the NEC, at 210.19(A) for branch circuits and 215.2(A) for feeders. They are recommendations, not enforceable limits, which is the single most misunderstood fact about voltage drop on a jobsite.

What does control the limit is the project specification, the equipment's listed voltage tolerance, and the adopted code edition with any local amendments. A spec can be tighter than 3 percent. A piece of equipment can list a minimum voltage that makes the rule of thumb irrelevant. When a contract calls out a number, that number wins over the informational note every time.

Treat a result within about half a percent of your target as a result that needs a second look, not a pass. Verify the routed length, the design current, the conductor material, and the temperature assumption before you accept it. The margin you think you have is the first thing that disappears when the load comes in higher than the schedule said it would.

Where the 5 percent budget actually goes

The 5 percent total is a budget shared across the whole path from the source to the load, and it gets spent in stages. The service or supply leg takes a cut, the feeder takes a cut, and the branch circuit takes the last cut. By the time the calculation reaches the receptacle or the equipment, the drop ahead of it has already eaten into the margin.

This is why a branch circuit that calculates at a clean 2.5 percent in isolation can still leave the equipment low. The feeder feeding its panel might already be at 2 percent before the branch starts. Add them and you are at 4.5 percent at the load, with no room for the day the load comes in heavy.

On a large building the practical move is to set a drop budget per stage early, hold the feeders tight so the branches have room, and check the worst-case path, which is the longest branch off the most distant panel. That path is where the total adds up, and it is the one the design should be sized around, not the average circuit.

Why is my voltage drop too high?

When a calculation or a measured reading comes back over target, the cause is almost always one of a short list, and they rank by how often they bite.

The run is longer than the plan said. Conduit goes up and over, around the structure, and through the pull boxes, and the routed length beats the straight-line plan distance by a wide margin on a real building. The current is higher than the design number, often because a continuous load was sized at its running value instead of 125 percent. The conductor material is aluminum where the calculation assumed copper, or the temperature in the space pushed resistance up. And sometimes the feeder ahead of the circuit was already eating margin nobody accounted for.

The fix is rarely exotic. You go up a conductor size, you shorten or re-route the run, or you move the panel closer to the load so the long leg carries less current. Upsizing for voltage drop is a design decision the NEC allows and expects, and when you do it, the equipment grounding conductor may need to grow in proportion. That last part is the step rookies skip.

Copper, aluminum, and conductor size

Copper has lower resistance than aluminum at the same size, so for the same run it drops less voltage. The trade carries aluminum anyway, because on larger feeders it costs less and weighs less, and you can still hit the target by going up a size or two. The calculation, not the metal, decides whether that works.

Going up one conductor size cuts resistance by roughly a third for each step, so it is the most direct lever you have. The catch is that everything connected to the conductor has to keep up. Lugs and terminations are rated for a conductor range and a temperature column. Upsize past what the lug accepts and you have created a new problem at the connection, which is where heat and failures concentrate anyway.

Mixing copper and aluminum resistance values in one calculation is a quiet, common error. Pull the resistance for the metal you are actually installing, from the same table, every time. If the conductor on the reel is aluminum, the copper number on your phone is not your number.

Upsizing the equipment grounding conductor

When you increase the phase conductors above the minimum size for any reason, including voltage drop, the equipment grounding conductor has to grow with them. The NEC handles this with a proportional-increase rule, commonly cited at 250.122(B): if the ungrounded conductors are upsized, the equipment grounding conductor is increased by the same ratio of circular-mil area. Skip it and the ground fault path is undersized for the conductors it is protecting.

This is the step that disappears on real jobs. The crew upsizes the feeder from #2 to 1/0 to fix a voltage-drop problem, pulls the same ground that suited the #2, and the inspector catches it, or worse, nobody does. The ground exists to carry fault current long enough to trip the breaker. Undersize it relative to the phase conductors and you have weakened the one path that has to work when everything else has gone wrong.

When you record a voltage-drop upsize, record the grounding conductor change in the same note. The two decisions belong together, and the person reading the record later needs to see that the ground was sized to the conductors that were actually installed.

A second example: aluminum feeder at the limit

Take the same 60 A load over 150 ft on a 208 V three-phase feeder, but pull aluminum #4 instead of copper. Aluminum #4 runs about 0.49 ohms per 1000 ft, so the calculation becomes 1.732 times 60 times 150 times 0.49 divided by 1000, which is about 7.6 V, or 3.7 percent of 208 V.

That fails the 3 percent branch target on the conductor alone, before any feeder drop ahead of it. To bring it back, go up to aluminum #2 at roughly 0.31 ohms per 1000 ft, and the same run drops to about 2.3 percent. This is the aluminum tax on a long run: you carry one or two sizes more than the copper equivalent to hold the same drop.

The decision is not copper versus aluminum on principle. It is cost and weight versus conductor size, run by run. On a short branch, aluminum rarely matters. On a long feeder, the upsize to hold the drop can erase the material savings, and that is exactly the calculation that should drive the call instead of shop habit.

Continuous loads: EV chargers, motors, and PoE

Voltage drop gets sharpest on continuous loads, because the current is not a peak that comes and goes. It sits on the conductor for hours and the heat builds. The NEC requires conductors and overcurrent devices for continuous loads to be sized at 125 percent of the continuous current, and that 125 percent is the current you should also be feeding into the voltage-drop calculation, not the bare running amps.

EV charging is treated as a continuous load under Article 625, and the runs are long because the chargers live at the far end of a parking field. A bank of Level 2 chargers on a long feeder is a voltage-drop problem first and a load-calculation problem second. Size the feeder for the drop over the routed distance, then confirm the ampacity, because the run length usually drives the conductor bigger than the ampacity table alone would.

Motors add inrush. A motor pulls several times its running current for the first moment of starting, and if the steady-state drop is already near the limit, the starting dip can drop out a contactor or stall the start. PoE is the same physics at low voltage: a powered device at the end of 90 m of Cat6 can fall below its class and reset, and the only fix is shorter cable, larger gauge, or a midspan.

Single-phase 120 V and 240 V runs

Most residential and small-commercial work is single-phase, and the formula uses the 2 factor for the conductor and its return. The voltage base changes the percent, which is what trips people up. The same volts dropped is a bigger percentage on a 120 V circuit than on a 240 V circuit, because the base is smaller.

A 120 V branch is where voltage drop shows up first, since it has the least margin to give. A long 120 V circuit to a garage receptacle, a well pump, or a run of exterior lighting hits 3 percent at a much shorter distance than a 240 V circuit carrying the same load. The fix is often to run the load at 240 V where the equipment allows it, which halves the current and cuts the drop roughly to a quarter for the same conductor.

When a homeowner reports lights dimming when the well pump or the AC starts, that is voltage drop plus motor inrush on a circuit that was sized for ampacity and never checked for distance.

DC runs: solar, battery, and control circuits

DC circuits drop voltage the same way, and the math is the single-phase form without the phase factor: 2 times current times one-way length times resistance, divided by 1000. There is no 1.732, because there is no phase angle. The 2 still accounts for the supply and return conductors.

DC voltage drop matters more than people expect on solar and battery work, where the source voltage can be low and the runs long. A low-voltage DC string or a battery feeder at 48 V has almost no percent margin, so a drop that would be trivial at 480 V is a real loss of usable power at 48 V. As battery storage and direct-DC distribution show up on more datacenter and site jobs, the DC drop calculation is becoming a routine check rather than a specialty one.

Control and instrumentation circuits have their own version. A 24 V control loop over a long run can drop enough that a relay or a sensor sits below its pickup voltage and behaves intermittently, which gets chased as a logic problem when it is really a wire problem.

When voltage drop drives the conductor, not ampacity

On short runs, the ampacity table sizes the conductor and voltage drop is an afterthought. On long runs, that flips. The conductor the load current alone would allow is too small to hold the drop over the distance, so voltage drop becomes the governing constraint and pushes the size up.

Knowing which case you are in saves time. A rough field rule: once a run gets into the low hundreds of feet at typical loads, start by checking the drop, because it will probably drive the size before ampacity does. Below that, size for ampacity and verify the drop as a check. The crossover depends on voltage, current, and the target, so it is a rule of thumb for where to start, not a substitute for running the number.

The expensive mistake is sizing the whole run for ampacity, pulling it, energizing it, and then discovering the equipment at the end runs low. Now the fix is a re-pull instead of a larger reel ordered up front. On long runs, check the drop before the conductor is on the order, not after it is in the pipe.

This is also where the estimate and the install part ways. The estimator priced the conductor the load table allowed, the field upsized it to hold the drop, and the difference becomes a change order or an erosion of margin nobody flagged. Catching the drop at takeoff, not at energizing, keeps that number where it belongs.

A field rule of thumb that holds up

Carry one number in your head and you can sanity-check a run on the spot: at 3 percent on a 208 V three-phase circuit, you have roughly 6.2 V to spend across the whole conductor. Watch the calculated drop climb toward that figure and you know you are running out of room before you finish the math.

The rule of thumb is for triage, not for the submittal. It tells you which runs to take seriously and which are obviously fine. Anything long, anything continuous, anything feeding sensitive equipment gets the real calculation with the routed length and the actual current. Everything short and ordinary gets the ampacity check and a glance. The skill is knowing which run is which before the conductor is on the truck.

Temperature, derating, and the resistance you actually have

Copper and aluminum both gain resistance as they warm, so the conductor running hot in a 40 degree C attic has more resistance than the same wire on the table in the shop. The NEC tables give resistance at a reference temperature, and the ambient correction and bundling derating factors in 310.15 adjust the ampacity you are allowed to use. Voltage drop tracks the same heat.

For most branch and feeder checks, the Table 8 DC value is close enough and the conservative move is to accept that your real drop runs a little higher than the table when the conductor is loaded and warm. On long runs and large conductors, AC resistance and reactance from Table 9 matter more, and the reactance term stops being something you can ignore.

The field lesson is simple. A calculation that just barely passes at room temperature does not have the margin it appears to have once the space heats up and the load comes on. Build in headroom on the runs that will run hot.

Power factor, reactance, and large feeders

The resistance-only formula is a good approximation, and on small conductors at near-unity power factor it is close to exact. On large conductors carrying inductive loads, it starts to understate the real drop, because the conductor's reactance adds to the resistance and the load's power factor changes how the two combine.

This is why NEC Chapter 9, Table 9 exists. It lists an effective AC impedance for conductors in raceway at a power factor of 0.85, which folds resistance and reactance into one number you can use the same way. On a big motor feeder or a long run of large conductors, using the Table 9 effective impedance instead of the Table 8 DC resistance gives a result closer to what you will measure.

The practical line is this: for branch circuits and modest feeders, Table 8 DC resistance is fine. For large conductors, long runs, and low-power-factor loads, reach for Table 9 and the effective impedance. If the load's power factor is well below 0.85, the drop can run higher than even Table 9 suggests, so build in margin.

Measuring voltage drop in the field vs calculating it

You calculate voltage drop before the pull to size the conductor. You measure it after energizing to confirm what you actually built. Both belong in the record, and they should agree within reason.

To measure, read voltage at the source and at the load with the circuit under real load, not at rest. An open-circuit reading at the far end tells you almost nothing, because voltage drop only exists when current flows. Put the load on, or apply a known test load, then take both readings as close to the same moment as you can and take the difference. A clamp meter confirms the current that goes with the reading, so you can compare it to the design current the calculation used.

When the measured drop beats the calculation badly, look for a loose or corroded termination before you doubt the wire. A bad connection drops voltage and makes heat right at the lug, and it shows up as a reading worse than the conductor alone should produce. That is the difference between a wiring problem and a workmanship problem, and the meter tells you which one you have.

Field checklist

What to document

Filed where the next person can find it, the voltage-drop calculation backs the conductor you pulled; lost in a truck, it backs nothing. The record is what answers the question six months out when a load runs low and the question is whether the conductor was ever right.

Capture the conductor material and size, the system voltage and phase, the design load current and whether the continuous adder was applied, the one-way routed length, the resistance source, the calculated volts dropped, the percent drop, the target you held to, who checked it, and the reason the calculation was triggered. If you upsized, write down what drove the size, because the next person will wonder why the feeder is bigger than the ampacity table demands.

Common mistakes

• Feeding round-trip length into a formula that already expects one-way length.
• Using the straight-line plan distance instead of the routed conductor length.
• Mixing copper and aluminum resistance values, or pulling resistance for the wrong size.
• Sizing the calculation on running amps when the load is continuous and needs 125 percent.
• Treating the 3 percent and 5 percent informational-note targets as enforceable code limits.
• Upsizing the phase conductors for voltage drop and leaving the equipment grounding conductor unchanged.
• Taking an open-circuit reading at the load and calling it a voltage-drop measurement.

Standards and references

The NEC, NFPA 70, is where the framework lives. Voltage drop itself appears as design guidance in informational notes, at 210.19(A) for branch circuits and 215.2(A) for feeders, recommending the 3 percent and 5 percent figures while stopping short of mandating them. The conductor ampacity that has to hold up alongside the drop comes from the ampacity tables, commonly Table 310.16 for the usual installation conditions, with the correction and adjustment factors in 310.15.

Conductor resistance values come from NEC Chapter 9, Table 8 for DC resistance and Table 9 for AC resistance and reactance in raceway. Continuous-load sizing at 125 percent comes from 210.19(A) and 215.2. EV supply equipment, a continuous load, falls under Article 625. The exact article and section numbers shift between code cycles, so confirm them against the edition the jurisdiction has actually adopted and any local amendments before you cite them on a submittal.

Equipment listings under UL and the manufacturer's instructions can impose a voltage tolerance tighter than any of these, and where they do, the listing governs. Cite the standard that controls the point, and let the project specification override the rule of thumb when it is stricter.

Units, terms, and conversions

Voltage drop goes by a few names and shows up in a few unit systems, so the same idea can read differently across a drawing set, a manufacturer sheet, and a spec.

Voltage drop is sometimes written Vd or called line loss. Conductor size is given in AWG for smaller conductors and kcmil (thousands of circular mils) for larger ones, while metric drawings use mm squared. Resistance is ohms per 1000 ft in the NEC tables and ohms per kilometer in metric sources. Percent drop is always referenced to the nominal source voltage of the circuit you are checking.

Vd / percent drop

Volts lost in the conductor, and that loss as a percentage of source voltage

AWG / kcmil

American Wire Gauge for smaller sizes, thousand circular mils for larger conductors

Ampacity

The current a conductor can carry continuously without exceeding its temperature rating

Continuous load

A load expected to run at maximum for 3 hours or more, sized at 125 percent

EGC

Equipment grounding conductor, which may need upsizing when phase conductors grow

Reactance

AC opposition from the conductor's magnetic field, significant on large or long runs

Related tools

Calculators and readiness checks for this work

Compare your options

FAQ

People also ask

• How far can PoE run?
• What size wire for a 48A charger?
• What is the difference between a feeder and a branch circuit?
• What is conductor ampacity?
• What is an electrical load calculation?