Concept

Capacity factor

Overview

The capacity factor is the ratio of a plant’s actual output over a period to what it would produce running at its full rated capacity the entire time — a number between 0 and 1 (often given as a percentage) that captures how fully the plant is used. It turns a nameplate rating into the production a model should actually expect.

Body

Definition. The capacity factor (CF) is

CF = actual output over a period / (nameplate capacity × that period)

— for an annual figure, the tonnes (or MWh) actually produced in a year divided by what the plant would make running at its nameplate rate every hour of the year (nameplate rate × 8,760 h). It is dimensionless, falls between 0 and 1, and is the fraction of the theoretical maximum a plant actually realizes.

What it bundles into one number. A single CF rolls together every reason a plant produces less than its rating: availability (the fraction of time it is online), the rate it runs at while online (turndown / utilization), the ramp of early operation, and — for a plant tied to a variable input — intermittency. The nameplate-vs-operating-capacity distinction is what the CF quantifies: it is that gap expressed as a ratio.

How it drives the economics. Annual production is nameplate × CF × time, so the CF sets the denominator under every per-unit figure. Costs incurred regardless of output — annualized total capex (via the capital recovery factor) and fixed opex — are spread over that production, which makes per-unit fixed cost scale as 1/CF: halve the CF and the fixed-cost contribution to levelized cost roughly doubles. Costs proportional to output are, by contrast, roughly CF-independent per unit. The CF is therefore a primary lever on the fixed-cost share of unit cost and largely irrelevant to the variable share.

Typical magnitudes. Continuous, dispatchable process plants run at high capacity factors — on the order of ~0.9 for a mature chemical plant — while plants tied to a fluctuating resource run far lower. Because the CF varies so widely by plant type, a value carried over from one type does not describe another.

Limits & typical error

• The two terms must share a basis. Actual and nameplate output have to be the same product, in the same units, with the same definition of “full capacity.” Computing the CF from a nameplate that is really an achievable average, or dividing an energy figure by a mass rating, returns a number that looks valid but means nothing.
• A period-average CF hides the time profile. The CF is an average over its period; two plants with the same annual CF can run completely differently within the year — one steady at 90%, another full-out then idle. For anything sensitive to when output occurs (intermittent-power matching, storage sizing, time-varying prices), the annual average conceals the behavior that matters.
• Per-unit fixed cost is hyperbolic in CF, so error amplifies at the low end. Because per-unit fixed cost ∝ 1/CF, the same absolute CF error moves unit cost far more at low CF: a 0.05 shift near CF 0.10 swings the per-unit fixed cost by tens of percent, while the same 0.05 near CF 0.90 moves it only a few percent. Low-CF estimates are intrinsically more sensitive to CF error than high-CF ones.
• A CF does not transfer across plant types. A ~0.9 borrowed from a continuous incumbent and applied to an intermittent-fed plant overstates output by a large factor, because a power-tied plant’s CF is set by its resource, not by process-plant norms. The realized CF travels with the resource and the operating mode, not with the process chemistry.
• Availability is not the capacity factor. Uptime is only one component. A plant online 95% of the year but running at 70% of rate while online has a CF of about 0.95 × 0.70 ≈ 0.66, not 0.95. Equating availability with the CF drops the rate term and overstates output.

See also

• Scale: nameplate vs. operating capacity — the design-vs-actual distinction the capacity factor quantifies as a ratio.
• Levelized cost — the per-unit cost whose fixed-cost share scales inversely with the capacity factor.
• Total capex — the capital that, once annualized, is spread over capacity-factor-determined output.
• Capital recovery factor (CRF) — how capex becomes the annual fixed charge the capacity factor then divides.
• Opex — fixed operating cost, spread over output like capital; variable opex is largely capacity-factor-neutral per unit.

Mini-example

A gray ammonia plant rated at ~1,000 t/day (a round anchor reused from the nameplate-vs-operating-capacity example) runs as a continuous, dispatchable process at a high capacity factor — take CF ≈ 0.90 . Its expected annual output is then

1,000 t/day × 365 days × 0.90 ≈ 330,000 t/yr

not the 365,000 t/yr the nameplate alone would suggest. Every per-tonne fixed cost — annualized total capex and fixed opex — is divided by ~330,000, and because the CF is high the gap between rated and realized output is small, so the fixed-cost denominator stays close to the nameplate ideal.

To show the edge, a separate one-line instance: a green ammonia plant whose hydrogen comes from electrolysis running directly on variable renewable power can carry the same ~1,000 t/day nameplate but a CF closer to ~0.3–0.5 set by the power resource , cutting annual output to roughly a third to a half and raising per-tonne fixed cost by the inverse — the regime where a small CF misestimate moves unit cost the most.

See also

• Scale: nameplate vs. operating capacity
• Levelized cost (e.g., $/t NH₃)
• Total capex (battery-limits → total plant)
• Capital recovery factor (CRF)
• Opex