Scope 3 Categories: The 15 Categories, Calculation Methods and Data Quality

For many companies, Scope 3 is the largest yet least understood part of the total footprint. In one of the examples in the GHG Protocol Corporate Standard, a logistics company found that 98 percent of its emissions in Sweden came from subcontracted haulage firms. What makes Scope 3 manageable is its 15-category structure and the calculation methods defined for each category. This guide covers both together. For a short reference to the categories, see our Scope 3 categories 1-15 quick reference; for where to begin, see Where to start with Scope 3.

Core rules

The binding requirements of the Scope 3 Standard (2011) are:

• All Scope 3 emissions are accounted for, and every exclusion is disclosed and justified.
• Each category is calculated according to the minimum boundary in Table 5.4 of the standard. The minimum boundary defines which activities must be included in each category; a company may go beyond it if it wishes.
• CO₂, CH₄, N₂O, HFCs, PFCs and SF₆ are accounted for where they are emitted in the value chain. The Scope 3 Calculation Guidance notes that a later amendment also made NF₃ mandatory.
• Biogenic CO₂ in the value chain is not added to the scopes but reported separately. Removals (for example biological carbon sequestration) are not included in Scope 3 and may be reported separately.
• Scope 3 emissions are reported by category.

The categories are split into two groups according to the direction of the financial transaction: upstream relates to purchased or acquired goods and services, downstream to sold goods and services. The categories are designed so that, for the same company, the same emissions are not counted in two categories.

Mind the time boundary: Scope 3 covers emissions associated with activities in the reporting year. In some categories (for example use of sold products and end-of-life treatment), the reported emissions have not yet occurred; they are the emissions that products sold in that year are expected to cause in the future.

The 15 categories: minimum boundary and calculation methods

The table below summarises the minimum boundaries in the Scope 3 Standard and the methods defined for each category in the Scope 3 Calculation Guidance (2013). In the guidance, methods are generally listed from the most specific to the most general.

Categories 1 and 2: four methods

Category 1 is the largest category and the one with the most method options. The guidance defines four methods:

• Supplier-specific method: Product-level cradle-to-gate GHG data are collected from the supplier.
• Hybrid method: Scope 1 and 2 data allocated to the product sold to you are obtained from the supplier; upstream emissions are calculated from the supplier's data on materials, fuel, electricity, transport and waste; gaps are filled with secondary data.
• Average-data method: The mass or another physical unit of the purchased good is multiplied by an industry-average emission factor.
• Spend-based method: The monetary value of the purchased good is multiplied by an industry-average emission factor per unit of spend.

The first two require collecting data from suppliers; the last two use secondary data. The formula follows the same logic in every case:

Emissions (tCO₂e) = Activity data (kg, units or spend) × Emission factor (tCO₂e per unit)

The guidance sets an important rule for Category 2: emissions from capital goods are not spread out through depreciation as in financial accounting, but reported in full in the year of acquisition. This category jumps in years of major investment, so context should be given in the report. Whether an item falls under Category 1 or Category 2 is determined by the company's own financial accounting practice.

Category 3: excluding combustion

The most common mistake in Category 3 is counting fuel combustion emissions a second time. According to the guidance, upstream emissions of fuels and electricity should be calculated with factors that include extraction, production and transport but exclude combustion, because combustion has already been counted in Scope 1 (fuels) or Scope 2 (electricity). For transmission and distribution losses, by contrast, the generation emissions of the lost energy, including combustion, are calculated. The guidance calls these "excluding combustion" factors upstream emission factors.

The general rule follows from this: in Scope 3, life-cycle factors are used for fuels and energy in the value chain; factors that contain only combustion emissions are for Scope 1 and 2.

Categories 4 and 9: transportation

There are three methods for transportation: the fuel-based method, based on the amount of fuel consumed; the distance-based method, based on the mass, distance and transport mode of the shipment; and the spend-based method, based on freight spend. The guidance notes that the GHG Protocol transport calculation tool uses two methods together, because CO₂ is better estimated from fuel and CH₄ and N₂O from distance. Site-specific and average-data methods are defined separately for storage. We cover how to collect tonne-km data in detail in our Category 4 and 9 article.

Category 11: potentially the largest category

Use of sold products can make up most of the total for companies that sell energy-consuming products or fuels. In the ABB case study in the GHG Protocol Corporate Standard, the use phase of a 4 kW low-voltage AC motor drive accounted for about 99 percent of total life-cycle emissions. The guidance describes separate calculations for three product types: products that directly consume energy (vehicles, engines, household appliances), fuels and feedstocks, and products that emit GHGs during use (refrigerants, fertilisers, fire extinguishers). The calculation is based on the number of lifetime uses per product, the number of units sold and emissions per use. Indirect use-phase emissions, such as washing clothes, are optional.

The Scope 3 Standard flags a trap here: because Category 11 counts lifetime emissions, a more durable product can increase reported emissions if the gain in lifetime outweighs the efficiency improvement. Reporting product lifetime and intensity metrics (for example emissions per hour of use) is therefore recommended as well.

Category 15: investments

For equity investments, the investment-specific method allocates the investee's Scope 1 and 2 emissions according to the share of investment; the average-data method estimates using revenue data and EEIO factors. For project finance, the share is calculated as the proportion of total project costs (equity plus debt). For financial institutions, this category is often almost the entire total.

Screening: order of magnitude first

The Scope 3 Standard says the most rigorous way to prioritise data collection effort is a screening:

1. Emissions are estimated for each Scope 3 activity using initial estimation methods (industry-average data, environmentally extended input-output (EEIO) data, proxy data or rough estimates).
2. Activities are ranked from largest to smallest by estimated emissions.

EEIO models allocate national emissions to product groups based on economic flows between sectors. The guidance notes that they cover the whole economy and are simple to apply, but that broad sector averages do not reflect process differences, that they assume a linear relationship between money and emissions, and that they struggle to show reduction efforts. EEIO is therefore the strongest tool for screening; in the final calculation it can fall short for large items.

Prioritising by spend is another option, but the standard gives a warning: spend and revenue may not correlate well with emissions. Just as there are high-value but low-emission items, there are low-value but high-emission ones. In one example in the standard, a company identified the items making up at least 80 percent of total spend, plus any item within the remaining 20 percent that individually exceeded 1 percent, and started its data collection there.

A practical example from the guidance shows mixed-method use: a company buying coffee beans from 100 suppliers collects primary data from the 10 suppliers that provide 85 percent of the volume and scales the result up to 100 percent; for sugar, which is small in total spend, it uses the average-data or spend-based method with secondary factors.

Relevance criteria

Size is not the only criterion. Table 6.1 of the Scope 3 Standard gives seven criteria for deciding whether an activity is relevant:

Exclusions

The standard aims for completeness but accepts that not everything may be calculable. The rules:

• No activity expected to contribute significantly should be excluded.
• Categories that do not apply (for example for a company with no leased assets or franchises) are reported as zero or not applicable.
• Every exclusion is disclosed and justified in the public report. In an example in the standard, a company excluded indirect purchases unrelated to production (such as office supplies), citing data constraints and an expected small contribution, and disclosed this.
• A manufacturer of intermediate products may exclude Categories 9, 10, 11 and 12 with justification if end use cannot reasonably be estimated, but cannot selectively exclude only some of them.

Data quality

Primary and secondary data

According to the guidance, primary data are data from specific activities in the company's value chain (supplier meter readings, invoices, engineering models, mass balances). Secondary data are industry averages, financial data, proxy data and other generic data. When requesting data from suppliers, the level most specific to the product is preferred: product level, then process or production line level, facility level, business unit level and, last, corporate level.

The guidance carries an important warning: specific data are not always accurate data. Supplier-specific and hybrid methods are more specific to the supplier, but the method the supplier uses to allocate its emissions to the product it sells you, and the reliability of its data sources, can increase uncertainty; supplier data can sometimes be less accurate than an industry average.

Five data quality indicators

The Scope 3 Standard recommends assessing data quality with five indicators:

In the standard's example scoring for temporal representativeness, less than 6 years of difference is "good", less than 10 years is "fair", and more than 10 years or unknown age is "poor". The standard also notes that this scoring involves subjectivity: some fuel factors have changed little over the years, so an older factor may not cause problems in practice.

The guidance describes data quality improvement as an iterative process and prioritises the point where two conditions meet: relatively low data quality and relatively high emissions. If an improvement in data quality changes emissions significantly, the base year should be recalculated; for details, see our base year guide.

SBTi and Scope 3

Under the SBTi Net-Zero Standard V2.0, a near-term Scope 3 target must cover every category within Categories 1-14 that on its own accounts for 5 percent or more of Scope 3 emissions. This means screening results directly determine target coverage: whether you calculate Category 4 at 4.8 percent or 5.2 percent can change the scope of the target. Improving data quality in categories close to the threshold should therefore be a priority. For the SBTi process, see our SBTi target-setting guide.

The mistakes we see most often in practice

• Counting combustion emissions a second time in Category 3
• Counting the same purchase twice across Categories 1 and 2
• Booking outbound logistics under Category 9 even though the company pays for it
• Calculating Category 11 on one year of use instead of the product lifetime
• Not matching inflation and currency year in spend-based factors
• Not disclosing exclusions in the report

For more on these and similar mistakes, see The 5 most common mistakes in Scope 3 data collection. If your customer asks for a product carbon footprint, that is a different exercise from a corporate inventory; see What to do when a customer requests a PCF.

At Alalëa, we start Scope 3 work with a spend-based screening, move to physical and supplier data for the items making up the first 80 percent, and document each category's method and data quality score in a methodology note. To build your Scope 3 inventory, see our carbon footprint service.