The GHG Protocol Corporate Value Chain (Scope 3) Accounting and Reporting Standard identifies fifteen distinct categories of indirect emissions that sit outside a company's owned or controlled operations. For most companies — particularly those in manufacturing, retail, professional services, or logistics — Scope 3 emissions represent 70 to 90 percent of their total GHG inventory. Yet the practical measurement challenge varies enormously across categories: some are straightforward to quantify, others require supplier engagement programs that take years to build. This guide works through all fifteen, with a focus on which calculation methods to use and where data collection is genuinely difficult.
Upstream Categories (1–8)
Category 1: Purchased Goods and Services
This is typically the largest and most complex Scope 3 category for product companies. It covers the cradle-to-gate emissions of all goods and services purchased in the reporting year. Three calculation methods are defined in the GHG Protocol Scope 3 Calculation Guidance: the spend-based method (apply an emission factor in kg CO₂e per monetary unit of spend from a database like Ecoinvent 3.10 or USEEIO); the average-data method (apply a mass-based emission factor, e.g., kg CO₂e per kg of steel); and the supplier-specific method (use actual emissions data provided by each supplier). Spend-based is the starting point for most organizations; supplier-specific data should be prioritized for the top 5–10 suppliers by spend that together typically account for 50–70 percent of Category 1 emissions.
Category 2: Capital Goods
Emissions from production of purchased capital goods — machinery, buildings, vehicles, IT equipment. The calculation method is the same as Category 1 (spend-based or mass-based), but capital goods are allocated over their full useful life in the year of purchase, not depreciated. Often lower materiality than Category 1 except in capital-intensive industries. Boundary note: if a company leases rather than owns its buildings and equipment, those may shift to Category 4 (upstream leased assets).
Category 3: Fuel and Energy-Related Activities Not in Scope 1 or 2
Covers upstream emissions from extraction, production, and transportation of fuels and energy consumed by the company but not counted in Scope 1 or 2. This includes transmission and distribution losses for electricity, and well-to-tank emissions for natural gas. The IEA Energy Statistics and DEFRA 2024 conversion factors both provide transmission loss coefficients and upstream fuel emission factors. This category is mandatory for companies using market-based Scope 2 accounting where the upstream T&D losses would otherwise be omitted.
Category 4: Upstream Transportation and Distribution
Third-party logistics emissions from getting products from suppliers to the company's facilities. The GLEC Framework v3 from the Smart Freight Centre is the standard methodology for freight transport calculations, providing mode-specific emission intensity values (e.g., tonne-km for road, sea, air freight). Data collection requires either logistics provider reports or transport activity data (shipment weight × distance × modal emission factor). This is a priority category for companies with complex international supply chains.
Category 5: Waste Generated in Operations
Waste treatment emissions — landfill, incineration, recycling, composting — from waste generated at the company's operated facilities. Calculation uses waste type, quantity, and treatment method, with emission factors from IPCC Waste sector guidance. Not typically a high-materiality category except for manufacturing companies generating large volumes of industrial waste.
Category 6: Business Travel
Emissions from employee travel by air, rail, and road for business purposes, excluding commuting. Air travel is calculated using flight distance, seating class, and DEFRA's 2024 aviation emission factors (which include radiative forcing uplift at 1.9×, though this is not required by all frameworks — GHG Protocol treats RF uplift as optional). Hotel stays are included under Category 6 under DEFRA methodology. Business travel data is usually available from expense systems or travel management companies.
Category 7: Employee Commuting
Emissions from employee travel between home and the regular workplace. Requires a commuting survey or assumption-based estimate using average commute distance and modal split data. Remote working adjustments are permissible and increasingly material: a home office energy calculation (using DEFRA home working factors) should offset some or all of a given employee's previous commute emissions depending on their work arrangement.
Category 8: Upstream Leased Assets
Operational emissions from assets leased by the company but not included in Scope 1 and 2 (typically where the lessee uses the equity share method for organizational boundary). Relevant primarily for companies with the operational control boundary that excludes landlord-controlled building energy, or for vehicle fleets on operating leases where the lessor owns the vehicle.
Downstream Categories (9–15)
Category 9: Downstream Transportation and Distribution
Emissions from transport of sold products from the reporting company's facilities onward to end customers or retailers — but only where the company does not pay for it (if the company pays, it's Category 4). Calculation methodology mirrors Category 4 using GLEC Framework. Often difficult to collect data for; spend-based approximation using outbound shipping costs as a proxy can be used as an interim measure.
Category 10: Processing of Sold Products
Emissions from processing of intermediate products by downstream companies. Relevant mainly for companies selling raw materials, ingredients, or semi-finished goods. If a company sells steel coils to manufacturers, the energy used in forming those coils into finished products is Category 10. This category is often immaterial for service companies or finished-goods retailers.
Category 11: Use of Sold Products
Lifetime operational emissions from the use of sold products by customers. This is the highest-materiality downstream category for many industries. For a company selling electric motors, fuel efficiency and electricity consumption over the product's useful life produce Category 11 emissions using IPCC AR6 GWP100 values (CH₄ = 27.9, N₂O = 273). For a software company, Category 11 would include the electricity consumed by customer hardware running the software — often quantified at a very small number per user.
Category 12: End-of-Life Treatment of Sold Products
Emissions from waste treatment of sold products at end of product life. Requires assumptions about product lifetime, mass, material composition, and end-of-life treatment mix (landfill vs. recycling vs. incineration rates by region). The GHG Protocol provides a sensitivity-based approach for estimating end-of-life rates when company-specific data is unavailable.
Category 13: Downstream Leased Assets
Emissions from assets owned by the reporting company but operated by a lessee. For a company that owns buildings and leases them out, the tenants' operational energy use appears in the owner's Category 13. This is a significant category for real estate companies but negligible for most manufacturing or service businesses.
Category 14: Franchises
Scope 1 and Scope 2 emissions of franchisee operations, if the reporting company is a franchisor. The franchisor discloses franchisee emissions at the Category 14 level. Relevant primarily for franchise businesses and not applicable to most B2B companies.
Category 15: Investments
For financial institutions and investors, Category 15 covers emissions associated with loans, equity investments, and project finance. Calculated using the Partnership for Carbon Accounting Financials (PCAF) standard, which provides attribution factors based on outstanding loan or equity shares. This is typically the dominant Scope 3 category for banks, insurance companies, and asset managers — and is the subject of increasing regulatory focus under ESRS E1-9 financial risk disclosures.
Prioritization: Where to Focus First
The GHG Protocol Scope 3 Standard explicitly permits companies to omit categories that are not relevant or are de minimis in size, provided the reasoning is disclosed. In practice, every company's materiality profile is different, but some patterns are consistent:
For a product manufacturer: Categories 1, 3, 4, and 11 are almost always material and should be quantified first. Category 1 (purchased goods) is where the biggest absolute emission reduction opportunities typically lie, and it is the category requiring the most intensive supplier engagement. Start with spend-based estimates, then migrate the top-spend suppliers to activity-based or supplier-specific data over successive reporting cycles.
For a services company: Categories 6 (business travel) and 7 (commuting) tend to dominate the upstream footprint, but are often already tracked through expense systems. Category 1 is still relevant — purchased IT hardware, office consumables, professional services subcontracting — but at lower absolute magnitude than for product companies. Category 11 may be significant depending on the software or service being sold.
We're not saying spend-based is an acceptable final state for a company with CSRD obligations — it is a starting point, not an endpoint. ESRS E1 expects companies to improve data quality over time, and an auditor reviewing a third-year CSRD report that still uses entirely spend-based factors for Category 1 will ask why supplier engagement has not produced better data. But for a first-year inventory, a defensible spend-based calculation with documented emission factor sources is materially more useful than no calculation at all.
The Data Architecture Problem
Consider a mid-size industrial components manufacturer with 40 tier-1 suppliers. The purchasing team tracks orders in an ERP system, but emission factors are not attached to any product category. Building Category 1 from scratch requires: extracting annual spend by supplier and spend category, mapping spend categories to USEEIO or Ecoinvent sector codes, applying the relevant emission intensity coefficient, and aggregating. This sounds mechanical — and it is — but the supplier-category mapping requires sustainability or procurement input to get right. Buying a "general industrial components" factor when most spend is in precision-machined aluminum is a methodological error that will surface in a DEFRA or Ecoinvent crosswalk review during assurance.
The goal in the first year is not perfect accuracy. It is documented accuracy — a methodology that explains what was measured, what factors were used, what was excluded and why, and what the uncertainty range is. That documentation is what external assurance providers evaluate under ISAE 3410, and it is what internal auditors need to sign off on ESRS E1 disclosures.