Cable sizing and the cable schedule for control panels

The short answer

A cable is sized by three checks, and the largest size wins: current carrying capacity after derating for ambient temperature, grouping and installation method; voltage drop over the actual route at design current; and short circuit withstand, from the adiabatic equation k squared S squared greater than or equal to I squared t. The cable schedule is the controlled document that records the answer for every cable on the job, and it is what site buys and pulls from.

Check one: current carrying capacity

The rule is a chain. Design current Ib must not exceed the protective device setting In, which must not exceed the cable's installed capacity Iz.

Iz is not the catalogue number. The catalogue gives a tabulated rating It for one reference condition, typically 30 degrees C in air or 20 degrees C in ground, one circuit, one installation method. Your installation is not that:

Iz = It x ambient factor x grouping factor x any soil, depth or enclosure factor

So the size you need is the smallest whose tabulated rating satisfies It greater than or equal to In divided by the product of those factors.

Take the factors from the applicable standard and the manufacturer, never from memory. In Indian practice that is IS 732, which follows IEC 60364-5-52 for installation methods and derating, and IS 3961 for recommended ratings, read with the cable maker's catalogue for the exact construction. Where the installation is not in the tables, a duct bank at a stated soil thermal resistivity for instance, the rating is calculated to IEC 60287 rather than looked up.

Check two: voltage drop

For a three-phase circuit:

Drop = root 3 x I x L x (R cos phi + X sin phi)

with L in km, R and X in ohms per km at operating temperature. Express it as a percentage of nominal. Projects typically work to about 3 per cent running from the distribution board and around 5 per cent overall at the point of utilisation; IS 732 and IEC 60364 Annex G give the recommended values. A motor circuit needs a second check at starting current, where power factor is low and current several times higher.

Check three: short circuit withstand

During a fault the conductor heats faster than it sheds heat, so the calculation is adiabatic:

k squared x S squared greater than or equal to I squared x t

which rearranges to S greater than or equal to I x root t / k, with S in square millimetres, I the RMS fault current in amps and t the disconnection time in seconds. The constant k depends on conductor material, insulation and permitted temperatures; IEC 60364-5-54 gives 143 for copper with XLPE, 94 for aluminium with XLPE, 115 for copper with PVC and 76 for aluminium with PVC.

Run the same check on the armour or protective conductor. On an earth fault the armour carries the current, and its k is lower and its area smaller, so it is often the binding constraint.

Worked example: a 30 kW motor at 80 m

Full load current. Take it from the nameplate. Without one, at 415 V with efficiency 0.92 and power factor 0.86:

I = 30 000 / (1.732 x 415 x 0.86 x 0.92) = 30 000 / 568.7 = 52.7 A, so design for 55 A.

Derating. Say the route is a tray at 45 degrees C with six loaded circuits touching in one layer. Read both factors from IS 732 or IEC 60364-5-52 for that exact reference method; here take 0.87 and 0.80. They are illustrative, not values to reuse.

It greater than or equal to 55 / (0.87 x 0.80) = 55 / 0.696 = 79 A

Select the smallest size whose tabulated rating in that method reaches 79 A. On a typical Indian LV job that lands on 3.5 core 25 sq mm aluminium XLPE armoured, 1.1 kV grade. Confirm against the manufacturer's table before committing.

Voltage drop over 80 m. For 25 sq mm aluminium, resistance at 20 degrees C is about 0.0282 x 1000 / 25 = 1.13 ohm/km; corrected to 90 degrees C at 0.004 per K, about 1.44 ohm/km. Reactance for a multicore LV cable is around 0.09 ohm/km; use the catalogue figure where you have it.

Drop = 1.732 x 55 x 0.08 x (1.44 x 0.86 + 0.09 x 0.51) = 7.62 x 1.284 = 9.8 V, or 2.4 per cent of 415 V. Inside a 3 per cent running limit.

Starting check. At a DOL start drawing six times full load, 330 A at power factor 0.3: 1.732 x 330 x 0.08 x (1.44 x 0.3 + 0.09 x 0.954) = 45.7 x 0.518 = 23.7 V, 5.7 per cent, so the motor still sees well above the 80 to 85 per cent usually required while starting.

Withstand check. Using the prospective fault current of, say, 35 kA for 0.1 s: S greater than or equal to 35 000 x 0.316 / 94 = 118 sq mm, which would condemn the cable. But a motor feeder is protected by a current-limiting device that never lets the prospective current through. Take the let-through I squared t from the manufacturer's energy limiting curve instead: at 0.3 million A squared s, S greater than or equal to root(300 000) / 94 = 5.8 sq mm, and 25 sq mm passes easily. The device data sheet decides this check, not the prospective current.

What a cable schedule is

A register, one row per cable, issued as a controlled document and revised as the job changes.

The electrical design engineer issues the power cable schedule; the control and instrumentation engineer issues the control and instrument schedule, because the screening and segregation rules differ. The client or consultant approves both; the contractor returns as-installed lengths.

Reconciling with the I/O list and the panel GA

Against the I/O list. Every signal must be traceable to a pair or core on the schedule. Multi-pair cables need their pair allocation recorded, usually on the loop drawing, or two engineers will use pair 3 for different signals. Spare cores belong in the schedule explicitly.

Against the panel general arrangement. The GA fixes gland plate size, entry face and number of entries. A schedule landing forty armoured cables on a plate with room for twenty-eight is discovered on site, at the worst moment. The GA also fixes internal segregation and whether there is depth below the gland plate for the cable's minimum bending radius, which the manufacturer states.

The mistakes people make

• Sizing on nameplate kW and stopping, with no derating for a hot tray full of cables.
• Using straight-line distance instead of routed length with risers, drops and a termination allowance.
• Checking voltage drop at running current only, then meeting the starting dip at commissioning.
• Skipping the armour withstand check, which fails before the phase conductors do.
• Letting the schedule drift from the I/O list after a late change.

What to learn next

Take one real feeder through all three checks with the actual tables in front of you, then draw its row in a schedule and trace the same cable onto a panel general arrangement. After that come protection device selection, where the let-through energy came from, and segregation between power, control and signal. Related reading: control panel wiring standards in India and how to read and produce a control panel GA drawing.