Concept

economic

Levelized cost (e.g., $/t product)

Levelized cost is the constant per-unit cost — dollars per tonne of product — that, charged on every unit produced over the plant’s life, exactly covers all capital and operating cost at the required return. It collapses an entire TEA into one comparable number, the standard measure of cost competitiveness.

What it is. It spreads every cost across every unit made, putting one-time capital and recurring operating cost on the same per-unit footing:

                 (CRF × total capex)  +  annual fixed opex  +  annual variable opex
levelized cost = ───────────────────────────────────────────────────────────────────
                                        annual output

The numerator combines the annualized capital charge (total capex × CRF) with annual opex. The denominator is actual output: nameplate × capacity factor × time, not the rating.

Its three shares. Capital, fixed-opex, and variable-opex (dominated by feedstock and energy). The first two are fixed costs spread over output, so they scale as 1/capacity factor; the third is roughly utilization-independent per unit. The mix is what gives a route its cost structure.

Why it’s the headline. One number per route makes competing routes comparable to each other and to the market price. It can be reported gross, or net of byproduct and policy credits and revenue. It means nothing without its basis — the system boundary, the capacity factor, and the CRF assumptions (cost of capital and life); two levelized costs compare only when those agree. It’s a cost of production, distinct from the market price — the gap is margin.

Limits & typical error

Non-comparable across mismatched bases — a gate-to-gate figure beside a cradle-to-gate one, or at a different discount rate, life, or capacity factor, compares unlike things; because the basis is so often unstated, this is the most common way “X is cheaper than Y” goes wrong (see system boundary).
One number hides the cost structure that sets its sensitivity — two routes at the same $/t can respond oppositely to a utilization or price shock; the single figure conceals the fixed/variable split.
It inherits the CRF assumptions and capacity factor, with a wide band — quoting it to several significant figures implies precision the underlying ±30–50% estimate can’t support, and dividing by nameplate instead of actual output understates it.
Gross-vs-net credit treatment changes the number, and it’s not a verdict — comparing a credit-netted figure to a gross one is apples-to-oranges; and it’s the cost to produce, not a market-clearing price or a return.

Levelized cost (e.g., $/t product)

Limits & typical error

See also