Voltage drop calculator to BS 7671
Voltage drop is the voltage lost along a cable under load: the cable's mV/A/m × the design current × the length ÷ 1000. Appendix 12 of BS 7671 limits it, from the origin of the installation, to 3% of the nominal voltage for lighting and 5% for everything else on a public supply, which is 6.9 V and 11.5 V at 230 V.
Cable size, current, length and what the circuit feeds. It works the drop out from the conductor's own resistance, judges it against Appendix 12, and says how far the cable could go, or what size would.
Appendix 12 of BS 7671 for the limits; the drop from BS EN 60228 conductor resistances at 70 °C, which is how the Appendix 4 mV/A/m figures are made. Checked against BS 7671:2018+A4:2026, updated 3 September 2026. Free, and nothing to sign in to.
What the load actually draws, not the device rating.
One way, board to load, along the route the cable takes.
The drop on the submain feeding this board. The limit is from the origin, so the two share it.
Voltage drop
0.0%
Inside the limit
0.0% of 230 V against a limit of 5% (11.5 V). The load sees about 230 V.
- mV/A/m for 2.5 mm², single phase 2 × 7.41 × 1.2
- 17.8 mV/A/m
- Drop along this circuit
- 0.00 V
- Limit for other on a public supply
- 5% = 11.5 V
- Longest run of 2.5 mm² inside the limit At this current, after anything lost upstream.
- 32 m
How is voltage drop worked out?
Current through a conductor's resistance loses volts, and the load at the end sees less than the supply gave. Appendix 12 of BS 7671 limits that loss, measured from the origin of the installation, to 3% for lighting and 5% for everything else on a public supply (6.9 V and 11.5 V at 230 V) and 6% and 8% on a private one.
The published tables give each cable a figure in millivolts per amp per metre. That figure is not arbitrary: it is the conductor's resistance at its 70 °C operating temperature, its BS EN 60228 value at 20 °C times 1.2, counted twice for a single-phase circuit, out along the line and back along the neutral, and √3 times for a balanced three-phase one. For 2.5 mm² that is 2 × 7.41 × 1.2 = 17.8 mV/A/m, which the table rounds to 18. The drop is then mV/A/m × Ib × length ÷ 1000.
Where the arithmetic stops
Up to 35 mm² the conductor's reactance is small against its resistance and this agrees with the tables to within rounding. Above that the reactance and the load's power factor start to matter and a purely resistive figure understates the drop, so the tool stops at 35 mm² rather than print a number it cannot stand behind. Those sizes want the tabulated r, x and z values.
Worked example
A 20 A load, a 4.6 kW heater say, at the end of 30 m of 2.5 mm² twin and earth, single phase, on a public supply.
- mV/A/m = 2 × 7.41 × 1.2 = 17.8.
- Drop = 17.8 × 20 × 30 ÷ 1000 = 10.7 V.
- 10.7 ÷ 230 = 4.6%, inside the 5% for a circuit that is not lighting.
- The longest run at 20 A that stays inside 11.5 V is 11.5 × 1000 ÷ (17.8 × 20) = 32 m. Two more metres and it would want 4 mm².
Were that a lighting circuit the limit is 3%, 6.9 V, and 2.5 mm² fails at 30 m, so 4 mm² is the smallest that passes.
Questions the trade asks
Appendix 12: 3% of the nominal voltage for lighting and 5% for everything else, measured from the origin of the installation, on a public low voltage supply, so 6.9 V and 11.5 V at 230 V. An installation on its own private supply is allowed 6% and 8%.
It is the conductor's resistance at 70 °C, its BS EN 60228 figure at 20 °C times 1.2, counted twice for a single-phase circuit, out along the line and back along the neutral, and √3 times for a balanced three-phase one. For 2.5 mm² that is 2 × 7.41 × 1.2 = 17.8 mV/A/m, which the published tables round to 18.
Yes. The percentage is from the origin of the installation, so a distribution circuit and the final circuit it feeds share one allowance between them. Put whatever has already been lost upstream in the box and the tool judges the total.
Above about 35 mm² the conductor's reactance is no longer small against its resistance and the load's power factor starts to matter, so a resistive calculation understates the drop. Those sizes need the tabulated resistive, reactive and impedance figures from Appendix 4, and the tool refuses to guess at them.
The design current of the load, Ib, not the rating of the protective device. A cooker circuit on a 32 A breaker that draws 25 A after diversity is a 25 A calculation. Where the load is genuinely unknown, a socket circuit say, sizing against the device is the safe assumption.
Next door
- kW to amps calculator, single and three phase (UK)Convert kW to amps and amps to kW at 230 V or 400 V, single or three phase, with power factor and the next standard device rating. Free, no sign-up.
- R1+R2 calculator: R1+R2 values and expected ZsR1+R2 for any copper cable and length, at 20 °C or 70 °C, and the earth fault loop impedance it gives with your Ze. Free, no sign-up.
- Maximum demand calculator with diversityAssess the maximum demand of a house after diversity: lighting, sockets, cooker, shower, EV charger, against a 60, 80 or 100 A supply. Free, no sign-up.
The figures go straight onto the certificate
In Pascal the maximum Zs is filled in for every circuit as you pick the device, the schedule of test results carries all 32 columns, and the observation library suggests the wording, code and regulation as you type. EICs, EICRs and minor works, from £25 a month.
See the certificate softwareA calculator applies a method to the numbers you give it. It does not know the installation in front of you, and it is no substitute for BS 7671, the guidance or the judgement of the person signing the certificate. Check anything you rely on.