Picking the right cable before the walls go up
How the two tests that decide a cable size actually work, what the controls on the page do, why a long run almost always needs a bigger cable than the load alone would suggest, and a walked-through example with real numbers.
Two tests, not one
A cable has to pass two separate checks before it can be used. The first is current capacity: can the copper carry the amps without getting dangerously hot? The second is voltage drop: does enough voltage still arrive at the far end? Short runs are nearly always decided by the first test. Long runs are nearly always decided by the second.
Most people only think about the first one. They look up how many amps a cable can handle, see that it covers their load, and stop there. That is fine when the socket is three metres from the board. When it is thirty metres away, the resistance of all that copper starts to matter, and the voltage at the far end can sag below what the regulations allow.
The calculator works through every standard cable size in order, from the smallest upward. For each one it checks both tests, applying the corrections for heat and grouping along the way. The smallest size that passes both is the answer, and it tells you which of the two tests was the one that decided it.
Telling it about the load
The load can go in as amps, watts or kilowatts. If you give it watts or kilowatts it works the current out using the voltage and, for alternating current, the power factor. A pure resistive load like a heater has a power factor of 1. A motor or older fluorescent fitting is closer to 0.85. If you are not sure, leave it at 1 and the result will be slightly on the safe side.
The supply can be single phase, three phase balanced, or direct current. Single phase is the normal domestic circuit. Three phase covers commercial feeds and some workshop machines. Direct current is for solar panels, battery banks and vehicle wiring. The voltage changes the current for a given wattage, so it matters.
Why the run length changes the answer
Copper is a very good conductor, but it is not perfect. Every metre of cable has a small resistance, and the current flowing through that resistance turns a small amount of power into heat. That heat is the voltage that goes missing.
The trick is that current capacity has nothing to do with length. A cable that can carry 32 amps when it is one metre long can carry 32 amps when it is a hundred metres long. But the voltage that reaches the far end drops with every extra metre. Double the distance and you double the drop.
The run length you enter is the one-way distance from the consumer unit to the equipment. The calculator doubles it internally, because current has to travel there and back. A 25 metre run means 50 metres of conductor in the circuit.
Installation method and why it matters so much
A cable clipped to a joist in open air can shed its heat freely. The same cable buried in loft insulation cannot. The insulation acts like a blanket, trapping the heat and forcing the cable to run hotter. A hotter cable can carry less current safely.
On the metric side the choices cover twin and earth cable in various situations, from clipped directly to a surface down to wrapped in insulation inside a stud wall. On the AWG side the temperature column picks the insulation rating, and the conductor choice switches between copper and aluminium. Aluminium carries less current per unit area but is lighter and cheaper, which is why it appears in larger feeder cables.
This setting moves the answer more than almost anything else on the page. Be honest about the worst part of the run. If most of the cable is clipped to a joist but one section passes through insulation, use the insulated setting for the whole thing. The cable does not know where the insulation ends.
The corrections for heat and grouping
The published capacity figures assume the air around the cable is at 30 degrees and that the cable is on its own. When either of those is not true, the capacity has to be reduced.
If six circuits are bundled together in one piece of conduit, each one can carry a good deal less than it could alone, because they are all warming each other up. The grouping correction can take the capacity down to barely half. It is the correction people skip most often, and it is the one that causes the most trouble in trunking that looked perfectly fine on paper.
The ambient temperature field adjusts for rooms that are hotter or cooler than the 30 degree baseline. A boiler room at 40 degrees reduces the capacity noticeably. An unheated garage in winter is a little kinder. Both corrections are applied before the current capacity test, so the number the cable has to beat is lower than the figure printed on the reel.
Walking through a real example
A shower rated at 9.5 kilowatts on a 230 volt single phase supply, with the cable run 18 metres from the consumer unit. The cable will be twin and earth clipped directly to the joists, on its own, in an ambient temperature of 30 degrees. The voltage drop allowed is set to 3 per cent.
The current works out to about 41 amps. At that load and that distance, 6 square millimetres can carry the current but fails on voltage drop. The answer comes back as 10 square millimetres, and the page says it was the drop that decided it. If the same shower were only 5 metres from the board, 6 square millimetres would pass both tests.
That gap between 6 and 10 is not a safety margin being generous. It is the resistance of 36 metres of copper refusing to be ignored. And the volts that went missing would have turned into heat in the floor void, every hour the shower was running, paid for at the same rate as the hot water.
Metric tables and AWG tables
The page offers two complete sets of cable data. The metric set uses square millimetre sizes and the BS 7671 Appendix 4 tables, which are the UK and most of Europe. The AWG set uses American Wire Gauge sizes and NEC Table 310.16, which is the United States and Canada.
Switching between them changes not just the sizes but the way the installation is described. The metric side has installation methods built around specific cable types. The AWG side has a temperature column and a conductor material choice instead, because the NEC tables are laid out differently.
One trap on the AWG side: the 75 degree column says 25 amps for 12 gauge copper, but the code does not allow you to protect it above 20. That small-conductor rule applies to 14, 12 and 10 gauge, and the calculator enforces it automatically so the answer stays honest.
What this does not replace
This is a planning aid. It tells you roughly what cable to buy and it helps you understand a quote from an electrician. It is not the wiring regulations and it cannot inspect your installation.
A real circuit also has to satisfy the protective device, the earth fault loop impedance, the disconnection time and the physical way everything is fixed in place. In most countries only a qualified and registered electrician can certify that work, and it is against the law for anyone else to do so. The cable size is just one line on a much longer list.
Use the number to have a more informed conversation with the person doing the work, not to skip that conversation entirely. If the answer surprises you, that is often a sign the run is longer than you thought or the installation conditions are tougher than you assumed.
Common missteps and how to avoid them
- Using the open-air capacity for a cable that passes through insulation, even briefly. The cable gets as hot as its worst section.
- Forgetting to count grouped circuits. Three cables touching in a channel all derate each other.
- Setting the load in watts but forgetting to lower the power factor for a motor. The current will be higher than expected.
- Entering the total cable length instead of the one-way run. The calculator already doubles it.
- Treating the percentage drop as trivial. A 3 per cent drop on 230 volts is about 7 volts, which is real power wasted as heat every hour the load runs.
- Comparing metric and AWG figures side by side without accounting for the different table structures. The two systems are not direct equivalents.
Everything stays on your device
Every calculation runs inside your browser. The numbers you type are not sent to a server, not stored in a database, and not visible to anyone else. You can disconnect from the network, reload the page, and it still works exactly the same way, because there was never anything on the other end to begin with.