Production capacity planning: working out what the plant can really make

Nameplate, demonstrated and effective capacity are different claims

Nameplate is what the equipment can do under ideal conditions with a single product and no interruptions; it belongs in a machine specification, not in a business case. Demonstrated capacity is what the plant has actually produced over a representative recent period, including its real setups, breakdowns, quality losses and absence. Effective capacity sits between them: demonstrated output adjusted for known, funded improvements. Confusing the three is how plants end up committing to volumes they cannot reach. Any capacity statement used in a commercial commitment should say which definition it uses and over what period the evidence was gathered.

Deriving demonstrated capacity from records rather than opinion

Take a period long enough to include the usual disruptions but recent enough to reflect current equipment and staffing, and calculate the output actually achieved per available hour on the resources that limit the plant. Exclude nothing, because the excluded events are the ones that recur. Then compare that figure with the best sustained week to see the gap between capability and typical performance; that gap is the improvement opportunity and it is usually larger than the gap that new equipment would close. Present both figures to whoever is deciding on investment, since a plant with a wide gap should be arguing about reliability, not about capital.

Capacity is only defined against a stated product mix

Output per hour varies by item, and changeover load varies by how the mix is sequenced, so a single capacity figure is meaningless without the mix it assumes. A plant quoted as capable of a certain weekly volume on long runs of one grade may deliver materially less when the order book shifts towards short runs of many grades. Model capacity for at least a baseline mix and a heavier-variety mix, and state the changeover assumption behind each. When commercial teams pursue small-batch business, this is the calculation that shows the hidden capacity cost of winning it.

Working up the ladder of capacity levers

The levers form a rough order of cost and reversibility: reduce losses on existing equipment, improve changeover practice, re-sequence to cut setup frequency, extend hours through overtime, add a shift, subcontract selected operations, then buy equipment or space. Most plants jump several rungs because losses are invisible and capital is visible. Before signing for a machine, require evidence that the loss and changeover rungs have been worked, expressed as recorded downtime and setup hours on the constraint. The exception is where the constraint is physically incapable of the required specification, in which case no amount of improvement will produce the capacity.

The lead time cost of running close to the ceiling

Utilisation and responsiveness pull against each other. As loading on a resource approaches its limit, queues in front of it grow faster than the loading does, and quoted lead times stretch even though output looks healthy. This is why a plant can be at record volume and simultaneously late for most customers. Decide deliberately how much headroom the constraint carries, and treat that headroom as a purchased service level rather than as waste. Plants that plan the constraint to full loading discover that every disturbance becomes a delivery failure, because there is nowhere to recover the lost hours.

Frequently asked questions

Should capacity be measured in hours, units or money?

Measure the constraint in hours, because hours are what you actually run out of and they are comparable across products. Convert to units only for a stated mix, and to money only when talking to finance about contribution. Plants that manage capacity in units alone get caught out when the mix shifts towards items that consume more machine time per unit, since the unit count stays flat while the plant runs out of time.

How often should capacity be recalculated?

Recalculate on a regular cycle tied to the planning review, and additionally whenever something structural changes: a new product with a different routing, a significant shift in order mix, equipment added or retired, or a change in shift pattern. The regular cycle catches drift, the triggers catch step changes. Leaving it to an annual budget exercise means the plant plans against a capability figure that stopped being true months earlier.

What if different departments disagree about the plant's capacity?

That disagreement is usually definitional rather than factual. Put the three definitions on one page with the evidence period for each, and ask each department which figure they are quoting and for what purpose. Sales generally wants demonstrated capacity with headroom, finance wants effective capacity for a business case, and engineering quotes nameplate. Once labelled, the argument narrows to a decision about headroom and improvement funding, which is a conversation that can actually conclude.