How voltage drop is checked
Every conductor has resistance, so current flowing through it produces a voltage drop along its length. If that drop is too large, equipment at the far end receives a lower voltage than intended — motors run hot, lighting dims, and some equipment simply won't start reliably. BS7671 Section 525 limits this to 3% of nominal voltage for lighting circuits and 5% for other circuits, when supplied directly from a public low-voltage network.
The calculation
BS7671 Appendix 4 publishes a millivolt-drop-per-amp-per-metre (mV/A/m) figure for every cable size, insulation type and configuration. Voltage drop in volts is:
V = (Length × Ib × mV/A/m) / 1000
divided by the number of parallel cable sets if more than one is run per phase. That figure is then expressed as a percentage of the nominal voltage and compared against the 3% or 5% limit.
Why longer runs need larger cable even when current rating is satisfied
Voltage drop scales with length, but current-carrying capacity doesn't. A cable sized purely on current rating for a 5m run can be badly undersized for the same load at 40m — the load hasn't changed, but the accumulated resistive drop has. This is the single most common reason a cable that "should" work on paper fails a full BS7671 check, and it's why sub-main and outbuilding circuits — garages, EV chargers, garden rooms — are so often limited by voltage drop rather than by how much current the cable can physically carry.
Worked example
The preset loaded in the calculator above — a 20A single-phase circuit, PVC twin-and-earth, method C, 45m from the consumer unit — returns a minimum compliant size of 4mm², at 4.30% voltage drop against the 5% limit (11.5V of 230V). 2.5mm² already carries 20A comfortably on current rating alone (its tabulated rating under method C is 27A), so it's the voltage drop over the 45m run that pushes the result up to 4mm². This is a non-lighting circuit at the 5% limit — a lighting circuit at the tighter 3% limit would need an even larger cable over the same run. Enter your own circuit above to see the exact figures.