Switchgear selection goes wrong in a predictable way: the board is chosen first, and the calculations are made to fit it. The results are usually visible two years later: a board with no room for the next production line, or one where the spare ways cannot be used because the busbar is already at its limit.
The sequence below is the one that avoids that. It is not complicated, and it takes less time than correcting the outcome.
Step 1: establish the demand
Connected load, demand factor, diversity and future expansion. The number you need is not the sum of the nameplates; it is the maximum current the board will realistically carry, plus a margin you can defend to your client.
Step 2: get the fault level
The prospective short-circuit current at the board, from the network study or from the transformer impedance. It sets two things: the short-time withstand rating of the busbar and the breaking capacity of every device downstream.
Where the calculation is not available yet, use a conservative assumption and write it on the drawing. An assumption that is visible can be corrected; an assumption that is silent becomes a defect.
Step 3: choose the assembly type
Panel type for main and sub-main distribution with moderate circuit counts. Modular, withdrawable assemblies where circuits must be maintained without shutting the board down, or where the number of outgoing ways is high.
This decision drives cost more than any other, and it is worth making consciously.
Step 4: set the enclosure
IP rating, material, finish and ventilation. Indoor electrical rooms are forgiving; plant rooms, wash-down areas and outdoor kiosks are not. Consider condensation as well as water: a sealed enclosure in a humid climate will produce its own moisture unless it is heated or ventilated.
Step 5: decide the form of separation
How far the operator is separated from live parts, and whether the busbar stays live while a feeder is worked on. The requirement is usually set by the plant's maintenance philosophy and by the standard the project follows.
Step 6: plan the metering
Incoming only, or per feeder? Local display, or data to a building or plant system? Metering is cheap at design stage and awkward to retrofit into a completed board.
Step 7: reserve the spare capacity
Spare ways, spare space and spare busbar capacity are the cheapest insurance in the project. Industrial plants change; boards that were sized exactly are the ones that get replaced early.
Step 8: confirm the documentation
Which standard the assembly is designed and tested to, and which documents you will receive. On a project with an independent engineer, the documentation is checked before the equipment is accepted, so it belongs in the specification, not in the correspondence.
What this looks like in practice
A mid-size plant with a 1600 A incoming, a fault level around 30 kA, twenty outgoing feeders and a maintenance team that works during production stoppages usually ends up with a modular assembly, IP54, metering at the incoming plus a few key feeders, and ten per cent spare ways.
A workshop extension with nine feeders and no expansion plan usually ends up with a panel-type assembly and a larger spare margin. Both are correct answers; the difference came from the questions in steps 1, 3 and 7.