Key takeaways
Short answer: Capacity and throughput answer two different questions. Capacity is what the plant could produce if everything ran perfectly — the theoretical ceiling. Throughput is what it actually produces, in good units, once downtime, slow cycles, and defects take their cut. The space between the two is lost potential, and it is almost always cheaper to recover than to buy. That gap is, in effect, what OEE measures. Before adding a line, find out how much hidden capacity the current one is leaking. For the framework, see OEE for manufacturing.
Capacity is the maximum output a process or plant could produce over a defined period under ideal conditions — every machine running at rated speed, no stoppages, no defects, for all the time it is scheduled to run. It is a planning number: it tells you the ceiling, what you are theoretically capable of. Capacity comes in flavours — design capacity (the absolute theoretical max) and effective capacity (the realistic max allowing for necessary changeovers, maintenance, and breaks) — but both are still idealised. Capacity is essential for planning and quoting, but it is not what you ship. Mistaking capacity for expected output is how plants overpromise and underdeliver.
Throughput is the actual rate at which a process delivers good output — real units, of acceptable quality, in real conditions. It is what you actually got, after downtime stole some hours, slow running shaved the rate, and defects removed some units from the count. Throughput is the honest operational number: it reflects reality, not potential. The crucial qualifier is good — throughput counts sellable output, so a unit that was made but scrapped does not count, even though it consumed capacity. Throughput is what fills orders and earns revenue, which is why it, not capacity, is the number that should drive operational decisions.
The difference between capacity and throughput is the most important number neither metric shows on its own. It is hidden, recoverable production — capacity you already pay for (the machines, the building, much of the labour) but do not convert into good output. Every hour of downtime, every minute of slow running, every scrapped unit widens the gap. The temptation, when throughput falls short of demand, is to add capacity: buy another machine, add a shift, build a line. But if the existing line is only converting a fraction of its capacity into throughput, that new capacity will leak at the same rate. Closing the gap is almost always faster and cheaper than buying more.
A line has the capacity to make 1,000 units per shift at rated speed. In reality it ships 600 good units. Management considers a second line to hit a 1,000-unit demand. But break down the gap first: the line lost the equivalent of 150 units to downtime, 150 to running below rated speed, and 100 to defects — 400 units of hidden capacity inside the existing asset. Recovering even half of that lifts throughput to 800 units with no capital at all. Buying a second line that also runs at 60% would cost a fortune to add the same 200 units the current line is already leaking. The gap, not the ceiling, is where the cheap output lives.
This is not an argument that capacity never matters. If your existing line is already converting most of its capacity into throughput — a high OEE — and demand still exceeds what it can produce, then you have a genuine capacity constraint and adding capacity is the right move. The discipline is sequence: measure how much of current capacity is actually reaching throughput before you spend. A plant running at 85% OEE that needs 30% more output needs more capacity; a plant running at 50% OEE that needs 30% more output almost certainly needs to fix the existing line first. Capacity decisions should follow the gap analysis, not precede it.
The capacity-to-throughput gap is essentially what OEE quantifies. Capacity assumes 100% availability, performance, and quality; throughput is what survives after each factor takes its loss. OEE breaks the gap into its three causes — time lost to stops (availability), speed lost to slow running (performance), and units lost to defects (quality) — so you know which loss to attack first. That decomposition into the six big losses is what turns a vague the line underperforms into a specific, prioritised list of recoverable output. OEE is the bridge between the theoretical ceiling and the real result.
Fabrico measures the throughput you actually achieve and, by tracking availability, performance, and quality losses against your rated capacity, makes the gap explicit and broken down by cause. Instead of guessing whether you need a new line or a better-run one, you see exactly how much hidden capacity is leaking and where. That turns the capacity-versus-throughput question from a capital debate into a data-driven decision. Book a demo to quantify the hidden capacity in your current lines.
Capacity is the maximum output a process could achieve under ideal conditions; throughput is the actual rate of good output it delivers in practice. Capacity is a theoretical ceiling; throughput is the real result, and the gap between them is lost potential.
They are related but not identical. Utilisation is the ratio of actual output to capacity, while throughput is the actual output rate itself. Throughput divided by capacity gives a utilisation figure, but throughput counts only good units.
Because real operations lose time to downtime, lose speed to slow running, and lose units to defects. Each reduces actual good output below the theoretical ceiling. OEE breaks the gap into exactly these three causes.
Not before checking the gap. If your existing line converts only a fraction of its capacity into throughput, new capacity will leak at the same rate. Fix the gap first; add capacity only when the current line already runs at high OEE and still cannot meet demand.
OEE essentially quantifies that gap. It measures how much of your theoretical capacity actually becomes good output, splitting the loss into availability, performance, and quality so you know which to address first.
Iniciar sesión
Blog
