85%
Ideal Utilization
3
Planning Horizons
Bottleneck
Sets System Capacity
Buffer
Absorbs Variability

What Is Capacity Planning?

Capacity planning is the process of determining whether your production system can meet customer demand — and what to do when it cannot. It answers the most fundamental question in operations: can we make what we need to make, when we need to make it?

Done well, capacity planning prevents overtime crises, missed deliveries, and expensive emergency actions. Done poorly (or not at all), it leads to chronic firefighting, bloated inventory, and broken customer promises.

Key Concepts

Rated vs. Demonstrated Capacity

TypeDefinitionUse
Rated (Theoretical)Maximum output if everything ran perfectly: no downtime, no changeovers, no defects, 100% speedUpper bound. Useful for investment decisions.
Demonstrated (Effective)What you actually produce, accounting for OEE losses: downtime, speed loss, scrapPlanning basis. This is the number you schedule against.

Never Plan to Rated Capacity

Planning to theoretical maximum is the #1 capacity planning mistake. If your machine is rated at 100 units/hour but your OEE is 65%, your effective capacity is 65 units/hour. Plan to demonstrated capacity, then work on improving OEE to close the gap. Use the capacity utilization calculator.

Bottleneck Capacity

Your system capacity equals the capacity of your bottleneck — the slowest step in the process. It does not matter if every other step can do 200 units/hour; if the bottleneck does 80, your system does 80. See Theory of Constraints.

Cut: 120/hr
→
Weld: 80/hr
→
Paint: 150/hr
→
Pack: 200/hr
System capacity = 80/hr (bottleneck). Improving any step except welding does not increase output.

Capacity Buffer

Running at 100% utilization sounds efficient but creates exponentially growing queue times. A 15-20% buffer absorbs normal variability (breakdowns, absenteeism, demand spikes) without missing deliveries. The math is clear: at 90% utilization, queue times are 3-5x longer than at 75%.

Four numbers called capacity 80/hr is the bottleneck's rate, not yet a plannable one. Put the page's own two remaining rules through it — × 65% OEE, × 85% utilisation — and what a planner may actually promise is 44.2/hr.

Every rate is the page's own. Cut 120, Weld 80, Paint 150 and Pack 200 units/hr come from the flow diagram under "Bottleneck Capacity"; 80/hr as the system capacity is that diagram's caption; 65% OEE is the "Never Plan to Rated Capacity" callout; 85% is the stat grid's ideal utilisation; the 15–20% buffer is the "Capacity Buffer" section. The arithmetic is the page's own formulas applied to the page's own bottleneck, which the page itself never does: Effective Capacity = Rated × OEE gives 80 × 0.65 = 52/hr demonstrated at the weld, and holding the buffer gives 52 × 0.85 = 44.2/hr plannable, or 52 × 0.80 = 41.6/hr at the other end of the range — a reserve of 7.8–10.4/hr. OEE loss at the constraint is 28/hr. The average, 137.5/hr, is (120 + 80 + 150 + 200) ÷ 4 — the number a spreadsheet reaches for, and the one the page warns against, since system capacity is the smallest step rather than the mean of them. The headroom above the constraint is (120 − 80) + (150 − 80) + (200 − 80) = 230/hr. As ratios, 44.2 is 55% of 80, 32% of 137.5 and 22% of 200.

The Three Planning Horizons

HorizonTimeframeDecisionsTools
Strategic1-3 yearsNew equipment, new lines, facility expansion, outsourcingDemand forecasting, financial modeling
Tactical1-6 monthsShift schedules, overtime planning, temporary labor, inventory buffersThroughput analysis, safety stock
OperationalDaily-weeklyJob scheduling, line balancing, changeover optimizationTakt time, hour-by-hour boards

Capacity Planning Process

Measure Current CapacityCalculate demonstrated capacity for every process step. Use OEE data, not nameplate ratings. Identify the bottleneck. This is your starting point.
Forecast DemandWork with sales and planning to understand demand by product, volume, and timing. Include seasonality, new product launches, and customer commitments.
Identify the GapDemand minus demonstrated capacity = gap. Positive gap means you need more capacity. Negative gap means you have excess (but keep a buffer).
Close the Gap (No Capital First)Before buying equipment: improve OEE on the bottleneck, reduce changeover time (SMED), rebalance the line, add shifts, reduce scrap. These options are faster and cheaper.
Capital Planning (If Needed)Only after exhausting operational improvements: new equipment, line duplication, outsourcing. Build the business case with demonstrated data, not theoretical capacity.

Quick Capacity Calculations

CalculationFormulaTool
Effective CapacityRated Capacity × OEE %OEE Calculator
Required Takt TimeAvailable Time ÷ DemandTakt Calculator
Operators NeededTotal Work Content ÷ Takt TimeLine Efficiency
Utilization %Actual Output ÷ Max Possible OutputUtilization Calculator
WIP RequiredThroughput × Lead TimeLittle's Law
✅ Good Capacity Planning
  • Plans to demonstrated (OEE-adjusted) capacity
  • Maintains 15-20% buffer for variability
  • Exhausts operational improvements before capital
  • Reviews capacity monthly against rolling forecast
❌ Capacity Planning Failures
  • Plans to rated/theoretical capacity
  • Runs at 95%+ utilization (queue time explosion)
  • Buys equipment before improving OEE
  • Only checks capacity during annual budget

🎯 Key Takeaway

Capacity planning is not a spreadsheet exercise done once a year — it is a continuous practice that connects demand to demonstrated capability. Always plan to OEE-adjusted capacity, always maintain a buffer, and always improve the bottleneck before buying new equipment. The cheapest capacity is the hidden capacity you already own — find it with OEE tracking, SMED, and line balancing.

Interactive Demo

Three work centres, each with its own available hours, efficiency and demand. Move any of them and watch where utilization lands — and which one turns red first.

⚡
Try It Yourself
Capacity Planning Simulator
▼
Adjust available hours, efficiency, and demand for each work center. Watch utilization bars change color: green (<85%), yellow (85-95%), red (>95%).
Machining
103%
-4h shortfall
AssemblyBottleneck
108%
-11h shortfall
Finishing
104%
-4h shortfall
Machining
160h
40h240h
85%
50%100%
140h
20h240h
Assembly
160h
40h240h
90%
50%100%
155h
20h240h
Finishing
120h
40h240h
80%
50%100%
100h
20h240h
105%
Avg Utilization
376h
Total Capacity
395h
Total Demand
-19h
Capacity Gap
Ready for the full knowledge check? Test your understanding with guided scenarios and data export.
PROTake the Pro Knowledge Check →
Free forever · Every feature included

Stop reading, start modeling

Model your process flow, run simulations, optimize staffing with TOC math, and test your knowledge with 107 interactive checks — all in one platform.

Open Workbench → Pro Knowledge Check

Take this to a room

The running order

For planners and operations managers. They should leave with demonstrated capacity rather than rated capacity as their planning basis, and a gap number.

7 beats · 11 min
  1. 1

    Rated capacity is a fiction you can plan against

    Planning to the nameplate is the single most common capacity mistake.

    • Rated: what the machine does if nothing ever goes wrong.
    • Demonstrated: what it actually produces, after downtime, speed loss and scrap.
    • A machine rated at 100 an hour with 65 per cent OEE gives you 65.

    Ask the room Which number is in our current plan?

  2. 2

    Three horizons, three sets of decisions

    Mixing them is why capacity meetings go in circles.

    • Strategic, one to three years: new lines, expansion, outsourcing.
    • Tactical, one to six months: shifts, overtime, temporary labour, buffers.
    • Operational, daily to weekly: scheduling, line balance, changeovers.
  3. 3

    The bottleneck is the capacity

    System capacity is not an average of the steps. It is the smallest one.

    • Adding capacity anywhere else changes nothing you can ship.
    • Measure demonstrated capacity at every step, then find the smallest.
    • That number is the plant's, whatever the other steps can do.
  4. 4

    Never plan to full

    Running at 95 per cent utilisation buys you queue time, not output.

    • Keep 15 to 20 per cent as buffer for variability.
    • The buffer is what lets you recover from a bad day without a cascade.
    • A plan with no buffer is a plan that only works on good days.

    Ask the room What utilisation is our current plan implicitly assuming?

  5. 5

    Close the gap without capital first

    There is a whole list of cheaper, faster options before anyone writes a business case.

    • Improve OEE on the bottleneck - that is capacity you already own.
    • Reduce changeover with SMED. Rebalance the line. Cut scrap.
    • Add a shift before you add a machine.
  6. 6

    Then, and only then, capital

    When you do build the case, build it on demonstrated data.

    • A business case on rated capacity will over-promise and under-deliver.
    • Show what operational improvement already recovered, and what is left.
    • New equipment at the wrong step buys nothing at all.
  7. 7

    The five calculations

    All of capacity planning fits on one card.

    • Effective capacity = rated x OEE.
    • Takt = available time / demand. Operators = work content / takt.
    • Utilisation = actual output / max possible.
    • WIP required = throughput x lead time.

    Ask the room Can we produce a demonstrated-capacity number for every step this month?