The depreciation instinct
Anyone who has worked in finance brings a reflex to carbon accounting: capital expenditure gets depreciated. A piece of manufacturing equipment that costs a large sum and runs for fifteen years doesn't hit the income statement in one shot — its cost is spread across its useful life. It's natural to assume Scope 3 Category 2 (Capital Goods) emissions should work the same way, with a machine's embedded carbon allocated year by year as it operates.
That assumption is wrong for the purpose of a corporate Scope 3 inventory, and understanding why clarifies a lot about how the GHG Protocol's Corporate Value Chain (Scope 3) Standard is actually structured.
What Category 2 covers, and how it's treated
Category 2 covers the cradle-to-gate emissions from the production of capital goods purchased or acquired by the reporting company in the reporting year — think machinery, vehicles, buildings, and large equipment that will be used for more than one accounting period. The standard calls for these emissions to be accounted for in full in the year the asset is purchased, not spread across its service life. There is no amortization schedule applied to the inventory itself.
This mirrors the treatment of Category 1 (Purchased Goods and Services), which also requires full cradle-to-gate emissions to be recognized in the year of purchase. The logic is consistency: both categories are measuring the embodied emissions of something the company bought, regardless of how long that thing sticks around afterward.
Category 1 versus Category 2: consumed versus built
The distinction between the two categories isn't about timing of emissions recognition — it's about what kind of good was purchased.
- Category 1 (Purchased Goods & Services) covers inputs that are consumed or transformed within the company's operations during the reporting period — raw materials, components, packaging, contracted services.
- Category 2 (Capital Goods) covers final products with an extended life that the company uses to run its operations rather than sell or consume — production equipment, IT hardware, vehicles, facilities.
Both get the same accounting treatment on the emissions side: full embedded emissions, booked in the year of acquisition. What differs is the nature of the asset, not the recognition timeline. A company replacing a fleet of delivery vans records the full cradle-to-gate emissions of those vans in Category 2 in the purchase year, exactly as it would record the full emissions of a batch of purchased steel in Category 1.
Why upfront recognition, not spreading
There are practical reasons the GHG Protocol avoids an amortization approach for the corporate inventory:
First, it keeps the inventory auditable against a single reference point — the transaction. Amortizing emissions over an asset's useful life would require every reporting company to track depreciation schedules, estimate remaining service life, account for early retirements or resales, and recalculate historical inventories whenever an asset's lifespan assumption changed. That's a heavy, assumption-laden burden layered on top of already uncertain emission factors.
Second, it avoids double counting and boundary drift. If a company spreads a machine's emissions over ten years, and three years in it changes its organizational boundary, sells a business unit, or the asset is scrapped early, the inventory has to be restated. Full recognition in the purchase year avoids that chain of dependencies.
Third, it matches the treatment already applied to Category 1 and keeps the two categories methodologically parallel — both report what was embodied in what was bought, in the year it was bought.
Where amortization-style thinking does belong
The useful-life allocation instinct isn't wrong everywhere — it's just the wrong tool for a corporate Scope 3 inventory. Amortizing a capital asset's embodied emissions over its operating life is a standard move in product-level life-cycle assessment and product carbon footprinting, where the question being answered is different: how much of this piece of equipment's footprint should be allocated to each unit of product it helps manufacture. That's a per-unit allocation exercise, not a corporate inventory exercise, and the two shouldn't be conflated when a company is also trying to meet reporting obligations under frameworks like CSRD or BRSR.
The practical effect: lumpy inventories
Because Category 2 emissions land entirely in the acquisition year, a company's Scope 3 footprint will show visible spikes in years of heavy capital investment — a new plant, a vehicle fleet refresh, a facility expansion — and relatively quiet years in between. That volatility is a feature of the accounting method, not a data quality problem, and it should be flagged explicitly when presenting multi-year trends so stakeholders don't misread a single large purchase as a structural increase in the value chain footprint.
It's also worth noting where Category 2 tends to sit on the Scope 3 data quality hierarchy — spend-based data (applying an emission factor to the dollar value of equipment purchased) at the weak end, average-data and hybrid approaches next, and supplier-specific primary data from equipment manufacturers at the strong end. Capital goods purchases are often infrequent and highly customized, which makes supplier-specific data harder to obtain than for recurring Category 1 purchases, so many companies lean on spend-based estimates here longer than they'd like.
Getting this category right matters disproportionately for capital-intensive sectors — manufacturing, heavy industry, infrastructure — where a single reporting year's equipment purchases can swing the total inventory more than a year of operational activity. Treating the purchase-year recognition rule as a known, expected feature of the methodology, rather than a reporting anomaly, is the first step to explaining it credibly to auditors, investors, and regulators.