Factory capacity utilization and investment decisions

Read the denominator before using the percentage

Capacity utilization compares actual production with an estimate of available capacity. The denominator is therefore as important as the output figure. A plant can show lower utilization after adding equipment even if it produces more goods. Conversely, retiring equipment can raise utilization without creating additional customer demand.

The Census Bureau's Quarterly Survey of Plant Capacity Utilization distinguishes actual, full and emergency production. It calculates industry rates from weighted production estimates reported by establishments. Its methodology also gathers work-pattern information and reasons for operating below full capacity. These distinctions are valuable because normal sustainable output and output under exceptional operating conditions answer different questions.

Inside a factory, first establish what the local measure includes. Is it based on a single staffed shift, multiple shifts, scheduled operating hours or a theoretical maximum? Does it count all products at the same assumed cycle time? A percentage without those definitions cannot be compared reliably across departments, let alone against a national series.

Use the same definitions when presenting an investment proposal and when checking the result afterward. Otherwise, a changed denominator can make the investment appear successful before any operating benefit has been demonstrated.

The national average is not a map of available capability

The December Federal Reserve release reported that September manufacturing output was unchanged from August and 1.5% higher than a year earlier. Utilization was 2.7 percentage points below the long-run manufacturing average. These are US aggregate estimates, incorporating the statistical vintage available with that release.

The same publication reported differences across industry groups. That matters for a supplier serving a particular production chain. A broad amount of spare capacity does not establish the availability of the machining envelope, clean environment, certification, tolerance or process capability required for its next contract.

Imagine a plant with several underused general-purpose machines but one fully booked inspection station. Purchasing another general-purpose machine would add nominal capacity while leaving the delivery constraint in place. Alternatively, a new process could enable work that the existing machines cannot perform at all. In that case, the proposal concerns capability as much as volume.

These are illustrative situations, not findings from the national release. Their purpose is to make the local question explicit: which customer requirement cannot be met with the resources already available, and what evidence shows that the proposed investment will resolve it?

Separate the causes of unused time

An idle hour can have several explanations. The plant might lack an order, an operator, a released drawing, an approved material or a functioning machine. Combining all those hours into a single utilization measure makes the number easy to report but difficult to act on.

The November Beige Book offers a related observation: some employers adjusted working hours to changes in business volume, and skilled roles remained difficult to fill in some places. This contact-based evidence reinforces the need to distinguish hours, staffing and equipment constraints. It does not quantify the spare capacity of an individual plant.

For an initial review, classify a representative period of lost time by observable cause. Avoid asking operators to infer the commercial reason for every interruption. They can record that a job lacked material; purchasing and planning can establish why. Preserve that distinction between observation and explanation.

Then compare remedies. If the constraint is job preparation, the relevant investment might be in tooling, fixtures, training or the release process. If a particular operation is genuinely overloaded, extra equipment may be appropriate. The objective is to identify the intervention that changes finished, acceptable output rather than simply increasing the number of assets on the floor.

Also examine whether work can be moved. Apparent spare capacity may require different tooling, programming, validation or customer approval before it becomes usable. Include those activities in the comparison. A machine that looks interchangeable in a capacity spreadsheet may not be interchangeable in the production plan.

A replacement project needs a different case from expansion

Manufacturers often combine several benefits in one proposal: more output, lower scrap, reduced maintenance and less operator attention. Those benefits may be real, but they should not be counted twice or assumed to occur at the same production volume.

NIST's earlier Investment Analysis Methods guide provides a useful framework for comparing cash inflows and outflows, discounting future amounts and examining uncertainty. It also addresses costs associated with organizational change and the capabilities required to implement a new technology. The guide is background methodology, not evidence that any particular machine will earn its projected return.

For a replacement, define the credible alternative. Continuing with the existing asset may require repairs, support or an eventual replacement anyway. For expansion, identify the additional work expected to use the capacity. For a capability investment, establish the qualification route and the demand that depends on successful qualification.

Keep these cases separate long enough to test them. A proposal justified by avoiding failures should stand up when sales growth is lower than hoped. A proposal justified by new sales needs evidence of those sales and the associated contribution after variable costs. It should not inherit the full value of a reliability benefit that is already counted elsewhere.

Test the proposal against a different operating year

A single forecast can make an uncertain investment look more precise than it is. Build alternatives that change the inputs most likely to affect the outcome: order volume, product mix, usable operating hours, implementation timing and recurring costs. Show what has been assumed rather than labeling one scenario as certain.

As an illustrative capacity calculation, suppose an operation has 2,000 available hours and each accepted job requires two hours. Its simple annual capacity is 1,000 jobs before additional allowances. Reducing the time to 1.5 hours gives about 1,333 jobs on the same basis. The difference matters commercially only if there is demand for the extra work and the rest of the production route can support it.

The example deliberately leaves out setup, scrap, downtime and product variation. A real proposal must include them. Its value is in exposing the link between technical improvement and saleable output. Faster processing alone does not specify revenue, cash savings or an acceptable investment price.

The Federal Reserve's manufacturing investment and capital-stock documentation provides another reminder about scope. Its November update used benchmark investment data through 2022 to support detailed industry estimates. Such series help explain the sector's capital base, but they cannot replace a current quotation or a factory's implementation budget.

Use a decision record that survives the purchase

Before approval, record the problem, the baseline, the intended change and the evidence required to confirm the benefit. Agree who measures accepted output, downtime, additional operating costs and customer acceptance. Specify the period over which the new process will be assessed.

Include implementation resources in the plan. A proposed delivery date for equipment is not automatically the date at which the factory can produce qualified parts at the expected rate. Installation, commissioning, process development and operator preparation need owners and realistic allowances.

Finally, set a review point for the assumptions. If the customer mix changes before commitment, the investment case should be revisited. That is a disciplined response to new evidence, not an admission that forecasting has failed.

The December data suggest spare capacity at the national level, but a useful factory decision is more specific. Invest when the proposed change addresses a demonstrated constraint or capability need and the economics remain credible under plausible alternatives. Treat the utilization percentage as the start of that examination.