The arithmetic that decides a station count: section size rather than class size, group size, experiment duration against the timetable slot, the rotation cycle, and the allowance for one rig being out of service.
The number that appears on most tenders is a guess dressed as a requirement. It is usually derived from the annual intake, which is the wrong number, and it is usually either double what the room can hold or half what the examination needs.
Start from the section, not the class. If sixty students are admitted a year and the timetable splits them into two laboratory sections of thirty, the room only ever holds thirty. Then apply one of two models.
Parallel model. Every group does the same experiment at the same time. Stations needed equals section size divided by group size. Thirty students in groups of four needs eight identical stations.
Rotation model. Every group does a different experiment and they rotate weekly. Stations needed equals the number of experiments in the cycle, one of each. Thirty students in groups of four needs eight different rigs and an eight week cycle to get everyone through everything.
Most engineering departments run the rotation because it is far cheaper. The cost of the rotation is timetable rigidity: the cycle length is fixed by the number of experiments, so eight experiments consumes eight teaching weeks with no room for a public holiday, a breakdown or a makeup session. Plan the cycle one or two experiments shorter than the number of usable weeks.
Group size is where budgets quietly damage courses. Three to four students per station is genuine participation: everyone touches the apparatus, takes a reading and gets a turn. Five to six is two students working and the rest watching, and the watchers write the same report from borrowed numbers.
So decide group size on pedagogical grounds first, then see what it costs. If the cost is unaffordable, the correct response is a longer rotation or a second session, not a group of eight. A department that runs groups of eight has, in practice, halved its practical curriculum without saying so.
Four situations override the simple calculation.
Sixty students admitted, two sections of thirty, groups of four, eight groups in the room, a three hour slot and roughly twelve usable teaching weeks. A workable pattern is eight to ten different experiments on rotation, of which two are duplicated so a group can be doubled when something fails, giving ten to twelve pieces of apparatus in the room.
Spend that budget unevenly and on purpose. Buy multiples of the cheap fundamentals, the ones every section needs in the first weeks of term: a Hooke's law unit, a jar test apparatus, a digital logic module, a beam deflection rig. Buy one each of the mid-cost rigs. Buy the expensive item once and timetable around it. Two mid-priced rigs almost always teach more than one premium rig, because the constraint in a teaching laboratory is student contact time, not instrument specification.
Then write the rotation before the tender is issued. If a plausible twelve week timetable cannot be drawn on one sheet of paper with the equipment list you are about to buy, the list is wrong, and it is far cheaper to find that out now than after delivery.