On board, heating does not determine the cross-section, but rather voltage drop. At 12 volts, three percent is only 0.36 volts — this limit is almost always reached before the thermal limit with typical cable lengths.
The Formula
FormulaVoltage drop in volts = (2 × cable length in meters × current in amperes × 0.0175) ÷ cross-section in square millimeters
The 0.0175 is the specific resistance of copper in ohms times square millimeters per meter. The factor of 2 is where most calculations fail: it refers to the single distance to the consumer, and because the current flows back and forth, it counts double.
Solved for the cross-section, which is the more common question in planning:
FormulaCross-section in square millimeters = (2 × length × current × 0.0175) ÷ permissible voltage drop in volts
The Limits
| Application | Permissible | at 12 V | at 24 V |
|---|---|---|---|
| Charging cables, electronics, navigation, lighting | 3 % | 0.36 V | 0.72 V |
| Non-critical consumers, short-term operation | 10 % | 1.2 V | 2.4 V |
Three percent is the value used for planning. Ten percent is an upper limit for consumers for which voltage fluctuations are less critical — not a target value.
Charging cables always belong in the first row. Whatever drops between the charging source and the battery is missing from the charging end voltage, and the battery bank will never be full — a mistake that shows up as a prematurely aged battery over the years.
Three Calculation Examples
Refrigerator compressor, 12 volts, 7.5 amps, 4 meters single run, target 3 percent.
Cross-section = (2 × 4 × 7.5 × 0.0175) ÷ 0.36 = 2.9 mm². 4 mm² is chosen as the next standard size.
Anchor winch, 12 volts, 80 amps, 12 meters single run, target 3 percent.
Cross-section = (2 × 12 × 80 × 0.0175) ÷ 0.36 = 93 mm². 95 mm² is chosen — a cable that is expensive and difficult to lay.
The same winch at 24 volts: 40 amps, target 0.72 volts.
Cross-section = (2 × 12 × 40 × 0.0175) ÷ 0.72 = 23 mm². 25 mm² is chosen. A quarter of the cross-section, the same consumer.
The third example is also the strongest argument in the system question: not the amperes determine the costs, but the product of current and length.
Four Ways to Reduce the Cross-Section
- Shorten the path. The most effective lever, because length enters linearly. Large consumers belong close to the battery — or the battery close to the consumer.
- Increase the voltage. Half the current with double the permissible drop results in a quarter cross-section.
- Shift the distribution. A sub-distributor near a group of consumers replaces many long individual cables with one thick short one and many thin short ones.
- Reduce the current. A more economical consumer has a double effect — on the balance and on the cross-section.
NoteThe sub-distributor is the most frequently overlooked lever. Instead of running six cables over twelve meters, one large cable goes there and six short ones from there — with appropriate fusing at both ends.
The Second Condition: Current Carrying Capacity
The calculated cross-section is a lower limit based on the voltage drop. It must also be able to carry the current thermally — and this capacity depends on the installation method. A bundle of cables in a closed conduit can carry less than a free-lying single conductor.
In practice, the voltage drop limit with typical lengths on board is usually above the thermal limit. For very short runs with high current — battery to main switch, main switch to starter — this is reversed: there, the current carrying capacity decides, not the calculation.
And regardless of both, the fuse protects the cable, not the device. It is selected for the cross-section, not the consumer.
What the Calculation Does Not Include
The formula describes the cable. It does not include the contact resistances at cable lugs, busbars, main switches, and fuse holders. A clean connection contributes almost nothing to this; a poor one can double the entire calculated value.
Therefore, every design requires a control measurement under load. If the measured value deviates significantly from the calculated one, it is not due to the formula, but to a connection point.
Don't Forget the Return Conductor
The factor of 2 in the formula assumes that the negative cable has the same cross-section as the positive cable. If it is made thinner — or omitted entirely and routed via the hull or frame — the calculation is no longer correct.
A ground connection via metal structures is problematic on board anyway: the resistance is not defined, changes with corrosion, and currents take paths that no one planned. Both conductors are routed, with the same cross-section.
Typical Mistakes
- Forgetting the factor of 2. Halves the result and leads to too small a cross-section.
- Planning with 10 percent. This is an upper limit for non-critical consumers, not a target value.
- Treating charging cables like consumers. What drops here is missing from the charging end voltage.
- Making the negative cable thinner. It carries the same current.
- Ignoring the installation method. In a bundle, the same cable can carry less current.
- Only calculating, never measuring. Contact resistances are not included in any formula.
