Choosing between similar factors

When several rows look right for your activity, choose the one whose unit, region, year window and boundary match your activity, then take one value for each methodology choice. Never add two candidates together: they are alternative answers to the same question, not parts of one answer.

This page gives the checks in the order to apply them, with real examples from factor search.

Two kinds of "similar"#

You meet two situations:

• Different factors for close activities: butane for Europe or for France, a van per km or per tonne-km. Each has its own factor page. You choose with the four checks below.
• Different values of one factor: the same activity published for several years, several GWP versions, or several electricity mixes. They live on one factor page, in Values by year and methodology. You choose along the axes further down.

In both cases the rule is the same: select one, never sum.

Check 1: the unit#

The factor's unit must be the unit of your quantity. A factor per kg cannot turn litres into emissions, and a factor per EUR cannot turn kWh into emissions. If your data is in money, you need a spend-based factor; if it is a physical quantity, an activity-based one. A money unit carries a currency as well: spend factors below covers that. See unit basis and units.

Watch out. Do not convert a unit with a density or a heating value you found elsewhere unless you state it. If factor search has the factor in your unit, take that one.

Check 2: the region#

Prefer the factor for the place where the activity happened. If none exists, take the smallest area that contains it, and say so. ADEME publishes butane for Europe at 0.269 and for France at 0.272 kgCO₂e/kWh; a site in France takes the France row.

Watch out. A rest-of-world region is not a wider stand-in for a country: it is the area it names minus the countries the publisher models separately. "Rest of World: Europe (EXIOBASE)" leaves Belgium out, so it is the other European countries, not a coarser Belgium. If a publisher models your country separately there is a row for it in the same library; if it does not, the rest-of-world row is the one built for you. See rest-of-world regions.

Some libraries state a region only by default: the row does not name one, so the library's home territory stands in. The factor page flags it with a tip on the region badge. See publisher default region.

Check 3: the year window#

Every value applies to a year window: "2019–", "–2018", "2019–2023" or "Any year". Take the value whose window covers your reporting year, not simply the newest. For a 2018 inventory, ADEME's butane for Europe is 0.26748, not the 0.269 that applies from 2019. See reference year.

Check 4: the boundary#

The boundary says which stages the value counts. "Combustion (tank-to-wheel)" counts burning only; "Well-to-use" adds producing and delivering the fuel. Two factors with different boundaries are not two estimates of the same thing. Pick the boundary your report asks for, and when you report direct and upstream emissions separately, use the parts from the Decomposition panel instead of the total. See system boundaries.

The value axes on one factor#

A factor can carry several values. Each one differs from the others along one or more of the axes below. For each axis, you either select one value, or report both side by side. In neither case do you add them.

To see them, open the factor page: Values by year and methodology lists the combinations, with a mark on the "publisher's default for this year". A free account opens all cells. In factor search, the Method filters (Approach, Grid mix, GWP method) narrow the list to one approach, grid mix or GWP version. A scenario is not one of those filters: EPD scenarios and other qualifiers shows how to ask for one.

Electricity: approach and grid mix#

AIB publishes one factor for grid electricity in France, per kWh, with four values:

For Scope 2 you report a location-based figure and a market-based figure, side by side. Within market-based you use one mix: the green or supplier value if your contract proves it, otherwise the residual mix. Adding the four gives a meaningless number. See calculation approach, grid mix and market-based vs location-based electricity.

Flights: radiative forcing#

DESNZ publishes a domestic flight in the UK, burning the fuel, per passenger-km, at 0.22928 kgCO₂e with radiative forcing (the default) and 0.13552 without. The first includes the extra warming of contrails and nitrogen oxides at altitude. Report the one your framework asks for, or both, labelled. See radiative forcing.

Spend factors: currency, basis and price level#

A spend factor is per one unit of one currency, at one price level, on one price basis. Three checks, in this order.

The currency. Read it off the Unit column, not off the library's country: EXIOBASE and ADEME publish per euro, USEEIO per dollar, Open CEDA per yen. Convert your spend to the factor's currency yourself, at a rate you write down, or pick a library that publishes in your currency.

The price basis. The same purchase has three prices, and the factor divides by one of them: basic (what the producer receives), producer (basic plus taxes on products) or purchaser (what you pay, margins included). EXIOBASE uses basic prices, Open CEDA producer prices, USEEIO purchaser prices. Factor search's tip says it: "The three bases are not comparable with one another." Your invoice is a purchaser price, so a basic or producer factor applied to invoice spend divides the same emissions by less money and overstates the result.

The price level. A euro of 2019 buys more than a euro of 2025, so a factor built on 2019 prices overstates 2025 spend. Factor search makes that adjustment for you: a spend factor is served once per year, and the row you are looking at is already expressed per unit of money of the year ticked in the Year filter. EXIOBASE animal products in France read 3.76 kgCO₂e per 2019 euro and 3.23 per 2025 euro, the same published figure brought forward.

So: pick the year of your spend in the Year filter, then multiply. Do not deflate again.

Watch out. The Currency year column is the publisher's own price level, the year the adjustment starts from. It is not the year your spend has to be in, and it does not change when you tick a different year.

Watch out. The price index runs to 2025. Spend from a later year has no adjusted row, and factor search does not extend the index by guessing. Report that spend on the latest year available and say so; never substitute the publisher's original figure or a zero.

See price basis, currency year and spend-based vs activity-based, which works the adjustment through year by year.

GWP: as published or normalised to AR6#

Each value is shown as the publisher published it, with its own GWP version: "GWP: AR6 (as published)". Libraries do not agree on a version, so factor search also keeps a second figure recalculated on AR6 over 100 years wherever that can be done exactly.

• For a DESNZ diesel van (Class III, per km, burning the fuel), the published AR5 value is 0.28046 kgCO₂e and the AR6 figure is 0.28051. The gap is small here because the value is almost all CO₂; it grows with the share of methane and nitrous oxide.
• When you add factors from several libraries in one total, use the AR6 figures, so that every line weighs the gases the same way. See GWP as published, normalised GWP and GWP.

Watch out. If a value has no AR6 figure, it could not be converted exactly. That is not a reason to put the published figure in its place: you would mix GWP versions in one total. Report that line on its own version and say so, or leave it out and say how many lines you left out.

Gases#

Never add the gases to the total: they are already in it. And a gas the publisher does not list is unknown, not zero. See By gas.

Supplier, EPD and data type#

Three labels describe a value rather than offer you a choice:

• Supplier. A supplier-specific value describes one supplier's product or one electricity contract. Use it only if you buy from that supplier. See supplier.
• EPD. An Environmental Product Declaration is one manufacturer's verified declaration for one product, not the category average. Use it when you know the exact product. See EPD.
• Data type. Whether the value is specific, generic or an average. See data type.

When you do not know the supplier or the product, prefer the average: in factor search, set "No EPDs" and "No named supplier".

When two candidates still tie#

If two factors pass every check, take the one from the library closest to your context (its home country, its sector), check the data quality rating, and write down why you chose it. Choosing a library helps with the first part. Consistency matters more than a small difference: use the same library for the same kind of activity across the inventory.

Next#

• Choosing a library: which library fits your activity.
• Cite a factor: record the choice you made.
• Read a factor page: where each of these checks shows.