Voltage Drop in Long Feeders: The Site Problem Nobody Calculates Properly
Design voltage drop is calculated at the source of the run and checked against a project target of 3 to 5%. Site voltage drop is measured from the far end of the run, at operating temperature, with all the parallel loads that were not modelled. That gap is where field problems appear.
The formulas and where they hide errors
The classic three-phase voltage drop equation is Vd = √3 × I × L × (R cosφ + X sinφ) / 1000, and it is what almost every design tool uses. The equation is correct. The inputs are where designers get sloppy.
Cable resistance R varies with temperature. Table values are usually at 20°C DC. At 90°C operating temperature, R is 27% higher. The manufacturer's mV/A/m tables account for operating temperature; the raw DC resistance from a datasheet does not.
Reactance X becomes non-negligible above 35 mm². Ignoring X on a large-conductor feeder underestimates the drop by 5 to 15%. On installations with parallel single-core cables in trefoil or flat arrangement, X depends on the spacing and formation.
Why site voltage drop is higher than design
Cable route length is measured door-to-door on the drawing; the actual installed route is longer because of vertical rises, offsets around obstructions, and pull-box provisions. A 10% length overrun is common.
Cable is not always the specified size. Contractors substitute cheaper equivalents on long runs where visual inspection is unlikely. Verify with cable schedules and random-sample identification.
Design load current assumes coincident diversified load. Site measurement often occurs during a peak event where the load exceeds design current. If diversity was assumed at 0.7 during design and the site test happens at 0.9 utilisation, the measured drop is proportionally larger.
Terminal connections add resistance. A poorly-crimped lug can add 0.5 to 1 mΩ that shows on the meter as an additional 0.2 to 0.5 V drop under load. Torque all terminations to specification and thermal-image at first energisation.
How to build in margin at design stage
Target 4% design drop when the code limit is 5%. The 1% margin covers length overrun and terminal resistance. On critical routes (fire pumps, elevator supply, generator emergency supply), target 3%.
Verify route length by walking it, not by scaling the drawing. Add 10 to 15% on top for pull-box provision, riser rise, and vertical offsets not shown on the schematic.
Split long runs at intermediate distribution points where possible. A 300 m run at 400 V starves the far end; splitting it into two 150 m runs from a mid-point sub-panel halves the drop.
Consider voltage step-up at source. On very long runs (site power distribution over 1 to 2 km), a step-up to 11 kV followed by step-down at the load reduces cable size dramatically and improves voltage regulation.
What to specify in the cable schedule
Column headings should include: cable ID, from-to, load kW, load current design, cable size proposed, cable size selected, one-way length, ampacity after derate, voltage drop at design current, voltage drop at 150% starting current (motor circuits), protective device size, and confirmed short-circuit withstand.
For motor circuits, always compute the voltage drop at motor locked-rotor current as well as at running current. A motor that can't start reliably is a design failure, and it usually shows up as an installation issue when the real cause is voltage drop.
Include the ambient temperature assumption on the schedule. For GCC projects, standard is 45°C ceiling-void ambient during summer peak. UK projects typically 30°C. Confirm with the M&E consultant before finalising cable sizes.
Field commissioning verification
Measure voltage at the incoming main and at each panel served under load. Compute the drop and compare to the design. If the drop exceeds design by more than 10%, investigate before handover.
Thermal-image every incoming breaker, sub-panel busbar, and major cable termination under full load. A termination with excess resistance shows as a hot spot 10 to 20°C above adjacent points.
Record the measured drop against each cable in the O&M manual. Baseline data at commissioning helps trace future problems back to a specific installation event.
Checklist
- Design voltage drop targeted at 3 to 4% (code limit less 1 to 2% margin)
- Route length verified by walk-through, not just drawing scale
- Ambient temperature confirmed with M&E consultant
- Motor circuits verified at both running and starting current
- Cable schedule includes drop at each check condition
- Intermediate distribution points considered on very long runs
- Terminations torqued to specification and thermal-imaged
- Baseline voltage drop recorded at commissioning
Common mistakes
- Using DC resistance at 20°C for AC voltage drop calculation
- Ignoring reactance on large conductors
- Measuring drawing length instead of installed route
- Not checking motor starting drop separately
- Skipping thermal imaging at commissioning
Frequently asked questions
What voltage drop tolerance should I allow beyond the code limit?
Around 1% margin for construction and length variance. If code is 5%, target 4% at design. If code is 3%, target 2 to 2.5%.
When is upstream voltage boost worth considering?
When the design drop pushes the cable up two or more sizes, or when the cost of the larger cable exceeds the cost of a transformer at the load side. Both conditions typically appear above 500 m of routing.
Do I test voltage drop or loop impedance at commissioning?
Both. Loop impedance confirms fault clearance time; voltage drop confirms operating performance. They measure different things.
Related calculators on this site
About this reference
Umar Dar is an MEP engineer with more than 15 years of experience in mechanical, electrical, and plumbing design and procurement across the Middle East and South Asia. He founded I Love MEP as a free resource for engineers, estimators, and procurement professionals.