Capital Goods (Scope 3 Category 2) — Definition and GHG Accounting Context
When a company builds a new factory, buys a fleet of trucks, or fits out a data centre, most of the carbon attached to that decision was already emitted — upstream, in the mills and plants that made the steel, concrete, and servers. Those emissions belong to the buyer’s carbon footprint, but they sit in a corner of the inventory that is easy to miss and easy to get wrong: Scope 3 Category 2.
Capital goods are the long-lived assets a company buys to operate, and Category 2 is the cradle-to-gate footprint of making them — booked in full the year they are acquired.
Capital goods are the long-lived assets a company acquires to operate — machinery, buildings, vehicles, and equipment. Scope 3 Category 2 is the cradle-to-gate emissions from producing them, counted in full in the year of acquisition rather than depreciated over the asset’s life. It differs from Category 1 only in that capital goods are capitalised assets, not expensed inputs.
Definition — What Category 2 Covers
Scope 3 Category 2 covers the greenhouse gas emissions from the production of the capital goods a company purchases in the reporting year. Capital goods are the final products an organisation acquires to provide a service, manufacture a product, or store, sell, and deliver goods — long-lived assets such as machinery, buildings, infrastructure, vehicles, tooling, and IT hardware. In financial terms they are the items a company capitalises and depreciates rather than expenses in the year of purchase.
The emissions counted are those of making the asset — its upstream, cradle-to-gate footprint, from raw-material extraction through manufacturing to the point it leaves the supplier’s gate. They are reported by the company that buys the asset, as part of its Scope 3 inventory, even though a different company did the emitting. This is the same principle that governs all upstream Scope 3 categories: you account for the emissions embodied in what you procure.
Category 2 is about the emissions of producing the asset, not of running it. Once the asset is in use, the fuel or electricity it consumes is the company’s own Scope 1 or Scope 2, not Category 2. Category 2 captures a one-off, cradle-to-gate quantity at the point of acquisition.
What Counts as a Capital Good (Category 1 vs Category 2)
The most common question is where Category 2 ends and Category 1 (Purchased Goods and Services) begins. The GHG Protocol draws the line by how the item is treated in financial accounting: capital goods are capitalised assets; purchased goods and services are expensed inputs. The carbon accounting is identical — both are cradle-to-gate — so the only real decision is which category an item lands in.
| Category 1 — Purchased Goods & Services | Category 2 — Capital Goods | |
|---|---|---|
| What it is | Inputs consumed or used up in the year | Long-lived assets used over many years |
| Financial treatment | Operating expense (OPEX) | Capital expenditure (CAPEX) — capitalised & depreciated |
| Examples | Raw materials, components, packaging, services, office supplies | Machinery, buildings, vehicles, IT hardware, tooling, infrastructure |
| Accounting boundary | Cradle-to-gate | Cradle-to-gate |
| When it’s counted | Year purchased | Year acquired — not depreciated |
Border cases follow the company’s own capitalisation policy: an item one firm expenses may be a capital asset at another, and it should be classified consistently with how it appears in the financial accounts. What matters for the inventory total is that every item is counted once, in exactly one category.
The Boundary: Cradle-to-Gate, and Where It Stops
Category 2 is a cradle-to-gate boundary: it includes raw-material extraction, upstream processing, transport between production stages, and the manufacture of the finished asset — everything up to the supplier’s factory gate. It deliberately stops there. Two things sit outside it and are frequently pulled in by mistake:
- The use phase. Energy consumed operating the asset is the buyer’s Scope 1 (fuel burnt on-site) or Scope 2 (purchased electricity) — never Category 2.
- The asset’s end of life. Disposal or recycling of the capital good, when it is eventually scrapped, is not part of Category 2 either. Any downstream credit for recovered materials is reported separately and is never netted against the acquisition footprint.
The Rule That Trips Everyone: No Depreciation
This is the single most important — and most often missed — rule of Category 2. Even though a capital good is financially depreciated over its useful life, its emissions are not. The GHG Protocol requires the full cradle-to-gate footprint to be reported in the year the asset is acquired, in one lump, with nothing spread across the years that follow.
A £5 million machine expected to last ten years is depreciated at £500,000 a year in the accounts — but its full cradle-to-gate emissions are booked to Category 2 in year one, not at one-tenth a year. Spreading (amortising) the emissions across the asset’s life understates the acquisition year and overstates the rest. A depreciation-style view can be shown as a supplementary disclosure, but the reported Category 2 total is always the full amount in the year of acquisition.
The direct consequence is that Category 2 is naturally lumpy: a year with a major capital programme — a new plant, a fleet renewal — will show a large Category 2 figure, and quiet years very little. That volatility is expected and correct, not an error to smooth away. When reporting trends, it is why capital goods are often shown separately or on a rolling basis alongside the raw annual number.
How Capital Goods Emissions Are Estimated
As with other upstream categories, the estimate is only as good as its data, and the GHG Protocol sets out a hierarchy from most to least specific. In practice a real inventory blends all three: supplier data for the few large assets, spend-based screening for the long tail.
| Method | Data used | Data quality | Best for |
|---|---|---|---|
| Supplier-specific | A product carbon footprint from the manufacturer (cradle-to-gate) | Highest | Major assets whose supplier reports a PCF |
| Average-data (asset / material) | Asset or material mass × a cradle-to-gate emission factor | Medium | Assets with a known bill of materials or type |
| Spend-based (EEIO) | Purchase price × an environmentally-extended input-output sector intensity | Lowest | Screening and the long tail of smaller assets |
Build an inventory across all three methods, with the currency and price-year normalisation the spend-based route needs, in the Scope 3 Category 2 Capital Goods Calculator, and see the full data hierarchy in the capital goods methodology.
Worked Micro-Example
A manufacturer buys one new CNC machine in the reporting year. It weighs 8,000 kg, mostly steel, and the average-data method gives it a cradle-to-gate emission factor of about 2.5 kgCO₂e per kg. (Figures illustrative.)
Cradle-to-gate emissions = 8,000 kg × 2.5 kgCO₂e/kg = 20,000 kgCO₂e = 20 tCO₂e
Booked to Category 2 in the acquisition year = the full 20 tCO₂e
Not spread over a 10-year life at 2 tCO₂e/year
Electricity to run the machine each year afterwards = the company’s Scope 2, reported separately
The whole 20 tCO₂e lands in one year’s Category 2, regardless of how long the machine will run. Its operating emissions are a different line of the inventory entirely — which is exactly why the no-depreciation rule and the use-phase boundary have to be applied together.
Units and CO₂e Basis
Capital goods emissions are reported in tonnes of CO₂e, like the rest of the inventory. Each greenhouse gas is converted to CO₂e by its global warming potential before being summed — methane at 29.8 and nitrous oxide at 273 times CO₂ over 100 years — so a Category 2 figure aggregates every gas embodied in making the asset into a single CO₂e total.
Common Confusions
- Depreciating the emissions. The full cradle-to-gate footprint is booked in the acquisition year, never amortised over the asset’s life.
- Confusing it with Category 1. Capital goods are capitalised assets; purchased goods and services are expensed inputs. Same cradle-to-gate accounting, different category.
- Including the use phase. Running the asset is the buyer’s Scope 1 / Scope 2, not Category 2.
- Miscounting a leased asset. An asset that is leased rather than bought belongs in the upstream leased-assets category, not Category 2.
- Netting end-of-life credits. Recycling or recovery benefits are reported separately and never subtracted from the acquisition footprint.
- Skipping the long tail. Smaller assets are easy to omit; a spend-based screen keeps them in without needing supplier data for each.
Frequently Asked Questions
Scope 3 Category 2 is the greenhouse gas emissions from producing the capital goods a company buys — the long-lived assets it capitalises and depreciates, such as machinery, buildings, vehicles, and IT hardware. The emissions counted are the asset’s cradle-to-gate footprint (raw materials through manufacturing to the factory gate), reported by the company that acquires the asset as part of its Scope 3 inventory. It excludes the emissions of running the asset, which are the buyer’s own Scope 1 or Scope 2.
Both cover the cradle-to-gate emissions of things a company procures; the difference is financial treatment. Category 1 (Purchased Goods and Services) covers inputs that are expensed and used up in the year — raw materials, components, services. Category 2 (Capital Goods) covers assets that are capitalised and depreciated over many years — machinery, buildings, vehicles. The carbon accounting method is the same for both; the item simply lands in whichever category matches how it appears in the accounts, and is counted once.
No. Although a capital good is financially depreciated over its useful life, its emissions are not. The GHG Protocol requires the full cradle-to-gate footprint to be reported in the year the asset is acquired, in one lump, with nothing spread across later years. Amortising the emissions understates the acquisition year and overstates the rest. This is why Category 2 is naturally lumpy — a big capital year shows a large figure and a quiet year very little. A depreciation-style view can be shown as a supplementary disclosure, but never as the reported total.
No. Category 2 captures only the cradle-to-gate emissions of making the asset. Once it is in use, the fuel it burns on-site is the buyer’s Scope 1 and the electricity it draws is the buyer’s Scope 2. Its eventual disposal is handled separately again. Keeping the acquisition footprint, the operating emissions, and end-of-life in their own lines is what prevents double-counting across the inventory.
Use the most specific data available, in this order: supplier-specific product carbon footprints for major assets; an average-data method (asset or material mass × a cradle-to-gate emission factor) where you know the asset type or bill of materials; and a spend-based method (purchase price × a sector emission intensity from an environmentally-extended input-output model) to screen the long tail. Most inventories blend all three. The Scope 3 Category 2 Capital Goods Calculator runs all three, with the full hierarchy in the capital goods methodology.