Capacity, Utilisation and OEE: Validating a Factory’s Real Output Potential

Capacity utilisation and OEE can support an investment thesis only when their definitions, source data and constraints are understood. Nameplate capacity is an engineering reference, not a promise of saleable output. Reported utilisation can be distorted by the denominator, while an impressive plant-wide average can hide a saturated bottleneck. Investors should build a capacity bridge by product and constraint, then connect it to volume, margin, working capital and capex assumptions.

Use four different capacity concepts

A disciplined review separates:

• Design or nameplate capacity: theoretical output under stated technical conditions.
• Demonstrated capacity: output actually achieved for a defined period, product and operating pattern.
• Sustainable capacity: output repeatable without unsafe practices, abnormal overtime, excessive breakdowns or deferred maintenance.
• Saleable capacity: conforming output that demand, customer approvals, logistics and working capital can absorb.

These are not interchangeable. A plant may demonstrate a short production peak by postponing maintenance or running a favourable product, yet be unable to sustain the forecast mix. Equally, low historical utilisation may reflect weak demand rather than a physical constraint.

Reconstruct the denominator

Ask precisely how management calculates utilisation. Possible denominators include calendar time, scheduled shift time, theoretical units, standard hours or a product-weighted equivalent. Check whether holidays, planned shutdowns, changeovers, trials and lack of orders are excluded. A changing denominator can create improvement without additional output.

Build an independent model from available days, shifts, hours per shift, planned stops, cycle rate and yield. For multiproduct lines, use standard minutes or constraint hours and test whether standards reflect the current routing and mix. Reconcile the result with production records, labour rosters, electricity consumption, raw-material issues and dispatches.

Interpret OEE carefully

Overall equipment effectiveness is commonly expressed as availability multiplied by performance and quality. It is valuable because it separates time lost to stops, speed lost while running and output lost to defects. It is not, by itself, a valuation metric or proof of spare capacity.

Test each component:

1. Availability: Does downtime include micro-stops, waiting for materials, changeovers and planned maintenance?
2. Performance: Is the ideal cycle time technically credible for the actual product, tooling and crew?
3. Quality: Are startup scrap, rework and downgraded output included, or only final rejects?

Review the raw timestamps and event codes behind dashboards. Manual overrides, uncategorised losses and inconsistent shift reporting can make a precise percentage unreliable. Compare automated data with operator logs and maintenance work orders.

ISO 22400 provides an industry-neutral framework for manufacturing KPIs, but diligence still needs to confirm the target's local definitions and data lineage. Benchmarking two plants is unsafe unless the calculation boundaries match.

Find the true system constraint

Capacity belongs to a value stream, not an isolated machine. Walk the routing from receiving through processing, inspection, packing, utilities, warehousing and dispatch. Identify the resource whose available constraint time limits saleable output at the forecast mix. The constraint may be a furnace, test laboratory, skilled operator, effluent-treatment plant, customer-approved tool or loading bay.

Then test interaction effects. Adding a shift may require maintenance cover, quality staff, canteen, transport and statutory permissions. Faster upstream output may increase work-in-progress rather than dispatches if downstream inspection is saturated. A new product can consume disproportionate setup time or constraint hours even when unit volume is small.

Validate evidence over time

Request at least enough history to capture seasonality, major maintenance and product changes. Useful records include:

• daily output by line, product and shift;
• scheduled and actual hours with downtime reasons;
• cycle-time and changeover distributions, not just averages;
• scrap, rework, first-pass yield and customer rejection data;
• maintenance shutdowns and failure logs;
• labour attendance, overtime and skill matrices;
• order book, dispatches and inventory movements.

Select peak, normal and weak periods for trace testing. If management claims an improvement, identify the physical or procedural change and confirm that the gain persists. Distinguish a repeatable control change from a temporary catch-up after shortages.

Translate capacity into the financial model

A capacity conclusion should change explicit assumptions rather than remain an operational appendix.

Growth consumes cash. Model inventory and receivables required for added sales, supplier terms, maintenance expense and commissioning losses. Separate debottlenecking, replacement and expansion capex, with realistic lead times and contingency. EBITDA uplift is not equivalent to free cash flow.

Use scenarios, not false precision

Develop downside, base and upside cases around constraint hours, OEE drivers, product mix and ramp-up. Attach evidence and confidence to each. An upside case should require identified actions, accountable owners, cost and time—not an assumed move to an external benchmark.

If a seller's forecast depends on unproven output, transaction responses may include a lower base valuation, staged investment, earn-out mechanics where appropriate, capex covenants or specific closing evidence. Legal and tax advisers should assess any contingent structure. The transaction advisory services page outlines related support.

Conclusion

Real factory potential is the output a complete system can repeatedly convert into accepted sales. By rebuilding the denominator, testing OEE data, locating the constraint and modelling the cash needed to grow, investors can replace a headline capacity claim with a defensible operating range.