Concept

Carbon intensity basis

Overview

A carbon intensity is the greenhouse-gas emissions attributed to a unit of product — kg CO₂-equivalent per kg of ammonia or hydrogen — and its basis is the accounting choices that make that number mean something: the boundary, the sources counted, the GWP horizon, and the co-product allocation. CI is to emissions what levelized cost is to money — a per-unit figure comparable only when its basis is stated.

Body

What it is. Total greenhouse-gas emissions over a defined boundary, as CO₂-equivalent, divided by units of product:

carbon intensity = Σ (consumption × emission factor) / units of product

Computed the same way as feedstock and energy cost, with an emission factor (kg CO₂e per MWh, per GJ, per tonne) in place of a price — the same physical consumption, valued in carbon instead of dollars.

The basis is the set of accounting choices — a CI number is defined by:

• the system boundary — gate-to-gate, cradle-to-gate, or wider;
• the emission sources counted — direct process, purchased energy, upstream supply chain (including fugitive methane for gas routes);
• the GWP horizon — usually 100-year, which sets how heavily short-lived gases like methane weigh; and
• the allocation between co-products.

Change any and the number changes, exactly as moving the boundary changes a levelized cost.

The dominant factor is often the energy source. For an electrified process, the electricity emission factor — grid-average, marginal, or contracted-renewable — frequently dominates CI and is the single most consequential basis choice. The product’s “color” (gray, blue, green) is itself a CI claim resting almost entirely on this.

What it feeds. CI is reported alongside levelized cost, and it sets eligibility and value for carbon and policy credits, whose thresholds are written in CI terms. A CI on one basis can’t be checked against a threshold defined on another.

Limits & typical error

• Non-comparable across boundaries and scopes — a gate-to-gate intensity omitting upstream emissions sits well below a cradle-to-gate one for the same plant; comparing them measures the basis, not the process (see system boundary).
• The electricity emission factor can swing the result by an order of magnitude — the same electrolytic plant is near-zero on dedicated renewables and as carbon-intensive as the fossil route on an average fossil grid; a favorable factor (annual-average vs. marginal, or renewables without hourly matching) understates CI dramatically.
• Co-product allocation can move emissions between products — splitting emissions by mass, energy, economic value, or a credit assigns materially different CI to the same product (the carbon analog of the byproduct-credit ambiguity).
• The GWP horizon changes methane-heavy footprints — GWP-20 weights methane far more than GWP-100; for routes with fugitive methane, CI depends visibly on which horizon is declared.
• Upstream emissions are easy to drop and usually understate — omitting feedstock production (upstream methane for gray/blue, grid losses for green) reports a CI real for the plant but below the delivered product’s; blue ammonia in particular hinges on capture rate and upstream leakage, both outside the fence.

Mini-example

Green ammonia’s CI is set almost entirely by the electricity basis. With ~10 MWh/t NH₃ (the energy intensity from the feedstock-and-energy example), CI is that consumption × the grid emission factor:

• on dedicated renewable power (~0 kg CO₂e/kWh), CI is near zero — the basis on which the plant is “green”;
• on an average fossil-leaning grid (~0.4 kg CO₂e/kWh), it’s ~10,000 kWh × 0.4 ≈ ~4 t CO₂e/t NH₃ — comparable to or worse than the conventional route.

Same plant, same chemistry; the “green” claim lives or dies on the electricity emission factor declared.

Edge case: gray ammonia carries a CI of ~1.6–2.4 t CO₂e/t NH₃ from reforming and fuel, and a blue variant’s CI isn’t a single number at all — it swings with the assumed capture rate and upstream methane leakage, so quoting one blue CI without that basis states a result the accounting hasn’t pinned down.

See also

• System boundary
• Feedstock & energy cost
• Energy intensity (per unit product)
• Carbon / policy credits
• Levelized cost (e.g., $/t NH₃)