What is the difference between rated and effective capacity?

Rated capacity is what a resource should produce at its normal rate and normal availability. Effective capacity is demonstrated capacity minus the planned losses you already schedule, such as maintenance, changeovers, breaks, and shift handovers. Effective capacity is the honest number to plan against; rated capacity usually overstates what a week will really deliver.

What is demonstrated capacity?

Demonstrated capacity is the best good-units rate a line has actually sustained over a real run under good conditions. It is measured from recent history, not calculated from a nameplate, so it cannot be argued away. It is the anchor you build a plannable capacity figure up from, rather than starting from the vendor's theoretical maximum.

How does OEE relate to capacity planning?

OEE is the multiplier that converts nameplate capacity into a plannable number, because it is the fraction of ideal output you actually convert to good product. Multiply the nameplate rate by OEE and you get close to demonstrated capacity. A line rated at 100 units an hour at 65% OEE should be planned as a 65-unit-an-hour line.

Why should you not plan against nameplate capacity?

The nameplate assumes the line never stops, never slows, and never makes scrap, which never happens. Planning against it overpromises output the plant has never achieved and guarantees a miss every period. Plan against effective capacity instead, built from demonstrated performance minus the planned and recurring losses you already know are coming.

How do you calculate a capacity number you can commit to?

Start from demonstrated capacity, the best rate the line has really held. Subtract planned downtime to get effective capacity, then apply the constraint's OEE to fold in recurring availability, performance, and quality losses. Check the result against takt time, and commit the master schedule to that effective number rather than the nameplate.

What is the cheapest way to add capacity?

Usually by recovering the effective capacity you already own rather than buying more. With average capacity utilization around 76 percent, much of the idle share is availability, performance, and quality loss that OEE measures. Attacking the largest of those losses often adds output at a fraction of the cost of new equipment.

Capacity Planning Metrics: Rated to Effective

The nameplate is a ceiling of hope. Rated, demonstrated, and effective capacity are the numbers you can actually plan against.

Capacity planning metrics are the numbers that turn a machine's nameplate into an honest promise about what a plant can actually produce. The core three are rated capacity (what it should do), demonstrated capacity (what it has done), and effective capacity (what it can do after planned losses).

Most planning goes wrong at the first number. Someone quotes the nameplate rate, multiplies it by the hours in a week, and promises output the plant has never once achieved. The gap between nameplate and reality is not fraud; it is changeovers, breakdowns, and slow running that the catalog never mentioned. This post defines the capacity metrics that matter, shows how OEE converts nameplate into a plannable number, and gives you a sequence for building a capacity figure you can actually commit to.

What are capacity planning metrics?

Capacity planning metrics are the layered measures of how much a resource can produce, from theoretical maximum down to what it realistically delivers. Each layer strips away another optimistic assumption, so the number you plan against is grounded in what the equipment and the schedule actually allow.

The point of the layers is honesty. Theoretical capacity assumes nothing ever stops; that is fiction. Demonstrated capacity is what you have really achieved; that is history. Effective capacity is demonstrated performance minus the planned losses you already know are coming. Plan against the theoretical number and you overpromise every week. Plan against the effective number and your schedule survives contact with the floor. Everything here ties back to your true capacity utilization actual output over what was really possible.

What is the difference between rated, demonstrated, and effective capacity?

Rated capacity is what the resource should produce at its normal rate; demonstrated capacity is the best it has actually sustained; effective capacity is demonstrated capacity minus the planned downtime you already schedule. They descend from optimistic to realistic, and the gaps between them are exactly the losses you can attack.

Metric	What it means	Where it comes from
Theoretical (design) capacity	Maximum output if nothing ever stopped, at full rate	Nameplate math; never achieved in practice
Rated capacity	Expected output at normal rate and normal availability	Design capacity adjusted for utilization and efficiency
Demonstrated capacity	Best output actually sustained over a real run	Measured from recent history, good conditions
Effective capacity	Achievable output after planned losses	Demonstrated minus changeovers, PM, breaks, mix
Actual output	What you truly produced	Measured; includes unplanned losses too

Demonstrated capacity is the anchor most plants skip. It is not a guess or a nameplate, it is the best rate the line has genuinely held over a sustained run, so it cannot be argued away. Build up from what you have proven, not down from what the vendor printed. A process capacity sheet is the tool that records this rate step by step.

Fig. 1, The capacity ladder: plan against effective capacity, not the nameplate at the top.

How does OEE turn nameplate capacity into a plannable number?

OEE is the bridge from nameplate to reality. Multiply theoretical capacity by OEE and you get a number close to demonstrated capacity, because OEE is exactly the fraction of ideal output you actually convert to good product. It bundles the three loss types, availability, performance, and quality, into one multiplier you can plan against.

Run the arithmetic and the honesty is obvious. A line rated at 100 units an hour with an OEE of 65% is a 65-unit-an-hour line for planning purposes, full stop. Schedule it at 100 and you will miss every week by the same 35%. This is why capacity planning and OEE are the same conversation: OEE quantifies the gap between the nameplate and what the plant delivers, and throughput is the good-units rate that comes out the other end. For the ceiling that includes unscheduled time, OEE vs. TEEP shows how much more capacity is hiding in the calendar.

One caution: use the constraint's OEE for planning, not a plant-wide average. A blended OEE across every machine flatters the number, because non-constraints with slack pull the average up while the bottleneck, the step that actually caps output, gets diluted. Plan the plant's capacity from the limiting step's loss profile, and let the non-constraints be as inefficient as they like without touching the promise. That is also why a single ugly week should not reset your capacity number: build the OEE multiplier from a representative stretch, not from the day everything went wrong or the day everything went right.

Fig. 2, OEE folds availability, performance, and quality into one multiplier on the nameplate rate.

Which capacity metrics should you track?

Track the layered capacity figures plus the loss and load metrics that explain them. Capacity alone tells you the ceiling; you also need utilization to see how much of it you use and the loss breakdown to see why the rest is gone.

Demonstrated and effective capacity in good units per hour, per line, your planning anchors.
OEE and its three parts the multiplier that converts nameplate to plannable, and the diagnosis of where capacity leaks.
Capacity utilization actual output over effective capacity, so you know whether the constraint is demand or the plant.
Takt time the customer's required rate, the yardstick your effective capacity has to beat.
Setup and changeover time a planned loss that eats effective capacity directly and is often the cheapest to recover.

How do you build a capacity number you can plan against?

Build it from proven performance, then subtract the losses you already know are coming. Do not start from the nameplate and hope; start from what the line has actually held and adjust for what the schedule demands.

Start with demonstrated capacity. Pull the best sustained good-units rate the line has really held over a solid run. This is your floor of proof, not the vendor's ceiling of hope.
Subtract planned downtime. Take out scheduled maintenance, planned changeovers, breaks, and shift handovers. What remains is effective capacity, the honest number.
Apply the loss profile with OEE. Use the constraint's OEE to fold in the availability, performance, and quality losses that recur even on good weeks.
Check it against takt. Compare effective capacity to the customer's takt time. If effective capacity cannot beat takt, no schedule will save you, you have a real capacity shortfall.
Attribute the biggest gaps. Rank the losses between demonstrated and effective. Changeovers, machine downtime or slow running, attack the largest first with the matching countermeasure.
Balance the line to the constraint. Use line balancing so effective capacity reflects the true limiting step, not an average that hides the bottleneck.
Feed it into the schedule. Commit the master schedule to effective capacity, not nameplate, so production planning promises what the plant can keep.

What do the numbers say about hidden capacity?

Most plants are sitting on capacity they have already paid for. Per the Federal Reserve's G.17 Industrial Production and Capacity Utilization release U.S. total-industry capacity utilization ran 76.2% in May 2026, roughly 3 percentage points below its 1972–2025 long-run average near 79%. That means about a quarter of installed capacity sat idle on average. Some of that is deliberate slack, but a large share is the availability, performance, and quality losses that OEE measures. The practical read: before you plan a capital expansion, plan to recover the effective capacity you already own. The cheapest capacity is almost always the capacity hiding inside your current OEE.

How do capacity metrics connect to demand?

Capacity metrics only matter next to demand. An effective capacity of 73 units an hour is neither good nor bad until you lay it against the load the schedule is asking for. That comparison, load versus available hours, week by week, is the job of capacity vs. demand planning and it is where a small, quiet overload turns into a missed ship date if nobody is watching.

The link runs both ways. Honest capacity metrics make the demand comparison trustworthy; a realistic master production schedule makes the capacity number worth calculating. Get the capacity numbers wrong and every downstream promise inherits the error. Get them right, built up from demonstrated performance, and the plan stops being a wish.

Where does real-time data fit in?

Demonstrated and effective capacity are only as good as the production data behind them. If your best-sustained rate is remembered rather than measured, your whole capacity ladder rests on a guess. That is the case for capturing counts and downtime reasons on the floor as they happen, so demonstrated capacity is a fact you can point to, not a number someone recalls.

Plants like CLS replaced paper production logs with real-time capture, so the demonstrated rate and the loss profile behind effective capacity come straight from the floor, current and defensible. Feed that into your planning and the schedule reflects the plant, not the catalog. To see what your losses are worth before you rebuild the plan, run your line through a free OEE calculator and compare the plannable number it gives you to the nameplate you have been promising against.