Double Counting Checker
For anyone worried that their Scope 3 counts something twice — and for anyone who has quietly left a category out to be safe. Seven questions separate the overlaps the GHG Protocol intends from the short list that are genuinely errors, and from the gap that fear of overlap usually creates instead.
Your Category 1 being your supplier’s Scope 1 is the design, not double counting. Real overlaps sit inside one inventory: leased assets, T&D losses, intra-group purchases. The commonest error is the opposite — Category 3 left out.
Seven questions, about a minute. Most of what people call double counting is the standard working as designed — your Category 1 is your supplier’s Scope 1. This separates that from the short list of overlaps that are genuinely errors, and from the gap that fear of overlap usually creates.
- Add Category 3a. It is not a second count of the same emission: Scope 1 covers combustion at your site, and 3a covers extracting, processing and transporting that fuel before it reached you. Different boundaries, same fuel, and both are required for full coverage — the fear of double counting is what causes the omission.
- Whether intra-group purchases are eliminated has not been checked. In a group that trades internally it is the easiest overlap to create and the hardest to see, because both entries are individually correct and only their combination is wrong.
What would change this verdict
- Switching consolidation approach moves the leasing answers. Under equity share a leased asset can legitimately appear in both Scope 1/2 and Category 8 with proportional allocation, where under operational control that same pattern is a double count.
- Switching Scope 2 to a “delivered” grid factor that already includes transmission losses would turn a separate Category 3 losses line into a double count. Standard generation-only factors never do.
Overlap register — each interaction, where you stand, and the rule
| Interaction | Your position | The rule |
|---|---|---|
| Category 3a — upstream fuel | Left out — Scope 1 assumed to cover itNOT a duplicate of Scope 1open | Scope 1 is combustion at site; 3a is extraction before it |
| Value-chain emissions | Reported in fullCorrect — the overlap is the designmet | Inter-entity overlap is intended, not double counting |
| Category 3 (c) — T&D losses | Reported separately in Category 3Reportedmet | Scope 2 grid factors are generation-only; losses go in Category 3 |
| Scope 2 method | Location-based onlySingle methodmet | Either method: Scope 2 factors exclude T&D losses |
| Category 8 — upstream leased | No leased assetsNot applicablemet | Under operational control a leased asset you control is Scope 1/2, not Cat 8 |
| Category 13 — downstream leased | Nothing leased outNot applicablemet | Whoever has operational control reports Scope 1/2; the other reports Cat 13 |
| Intra-group purchases | Not checkedNot establishedopen | A consolidated entity’s Scope 1 is already in the group total |
Why this verdict: Something has been left out of the inventory to avoid an overlap that the standard intends. Value-chain accounting means one entity’s Scope 3 IS another entity’s Scope 1, and that inter-entity overlap is the design rather than an error — it exists so decarbonisation pressure sits with everyone who can act on it rather than with the producer alone. Category 3a is the same misreading in a narrower form: Scope 1 covers combustion at your own site and 3a covers extracting, processing and transporting the fuel before it arrived, so they are different boundaries on the same fuel and both are required. Category 3 is the category most commonly missing from inventories that calculate Scope 1 and 2 correctly, which is why it is a standing verification finding — and the reason it goes missing is precisely the instinct that produced this verdict.
This checks overlap only. Each emission appearing once says nothing about whether every category that should be present is present, or whether the data behind it is good — completeness and quality are separate tests. Nothing here is quantified either: what an upstream component is worth is a calculation, and this page only says whether it belongs and where.
The Corporate Standard is under revision and the treatment of value-chain boundaries is part of it. We will email you when something this checker applies changes — not otherwise.
This checker applies the double-counting mechanics of the GHG Protocol Scope 3 Standard as our reviewed standards page records them. It does not read your inventory, it does not quantify anything, and it covers the interactions the standard identifies as genuine risks — it tells you which of your overlaps are errors and which are the design.
What this verdict means
Five positions, ordered by how wrong the belief behind each is. That puts the missing category above the genuine duplicate, because the person who deleted something to be safe believes their inventory is the more rigorous for it.
Understated
Something is missing because it looked like a duplicate. The overlap you avoided was not one, and removing it created a gap rather than preventing an error.
Counted twice
A real overlap, inside one inventory, where the standard expects each emission once. Both entries usually look right read on their own — only the combination is wrong.
Interaction unresolved
One of the interactions that genuinely can double count has not been established. It turns on a boundary decision rather than on arithmetic, so it can be either until somebody decides.
Treatment undeclared
Each emission appears once, and the reasoning that put it there has not travelled with it. A reviewer cannot tell a considered choice from a coincidence.
No overlap found
Upstream categories present, leasing treatments on one side each, nothing excluded for fear of duplicating a supplier. A statement about overlap, not about completeness.
The overlap that is the design
The Scope 3 Standard is built around value-chain accounting, and value-chain accounting overlaps between companies deliberately. Your Category 1 purchased-goods figure is your supplier’s Scope 1 figure. Their Scope 3 downstream is your Scope 1. The same tonne appears in several inventories at once, and that is the mechanism working.
It exists because a single-count system concentrates all decarbonisation pressure on the producer, who is frequently the party least able to change what is produced. Overlapping it puts the same emission in front of everyone who can act on it — the buyer specifying the material, the financier funding the plant, the manufacturer running it.
Inter-entity overlap is not double counting and needs no correction. Genuine double counting is a much narrower thing: the same emission appearing twice inside one inventory. Those are the only cases worth hunting, and there are five of them.
The practical consequence is that you should never remove something from your own inventory on the grounds that a supplier also reports it. That does not prevent a double count; it creates a hole in your own figure while leaving theirs untouched.
Category 3, and why it goes missing
Category 3 — fuel- and energy-related activities — is the most universally applicable Scope 3 category. It applies to every organisation that burns any fuel or buys any electricity. It is also the category most commonly omitted from inventories that otherwise calculate Scope 1 and Scope 2 correctly, and the omission is consistently flagged as a verification finding.
It goes missing for a specific and understandable reason: 3a looks like counting the same fuel twice.
| Component | What it covers | Boundary |
|---|---|---|
| Scope 1 combustion | Burning the fuel at your own site. | Tank-to-wheel — from the moment it arrives. |
| Category 3a | Extracting, processing and transporting that same fuel before it reached you. | Well-to-tank — everything up to the moment it arrives. |
They are the same fuel and different boundaries, which is why both are required and neither duplicates the other. For UK natural gas billed on gross calorific value, the upstream component is 0.03021 kg CO₂e/kWh [GreenCalculus fuels.gbr.natural_gas.wtt.kwh_gcv · DEFRA 2026 'WTT- fuels'!D41 · v2026.284] against a combustion factor of 0.18231 kg CO₂e/kWh [GreenCalculus fuels.gbr.natural_gas.kwh_gcv · DEFRA 2026 'Fuels'!D42] — a material addition, not a rounding difference, and one that disappears entirely if 3a is skipped.
Transmission and distribution losses go missing the same way. They are Category 3 activity (c), and it is easy to assume the grid factor already contains them. It does not. The GHG Protocol Scope 2 Guidance says both Scope 2 methods use “generation-only emission factors” and that “the emission factors do not include T&D losses”, so the losses belong in a separate Category 3 line whether you report location-based, market-based or both. The one genuine double count is a separate losses line on top of a “delivered” Scope 2 factor that already includes them; the fix is to use a generation factor for Scope 2.
The overlaps that are real
Five interactions inside a single inventory can genuinely double count. What they share is that each turns on a boundary decision rather than on arithmetic — which is why the same figure can be correct under one consolidation approach and duplicated under another.
| Interaction | Double counting? | The rule |
|---|---|---|
| Cat 3a (upstream fuel) + Scope 1 combustion | No | Different boundaries on the same fuel. Both required. |
| Cat 1 (purchased goods) + supplier’s Scope 1 | No — intentional | Inter-entity overlap is the point of value-chain accounting. |
| Cat 4 (upstream transport) + Cat 9 (downstream) | No | Different ownership — inbound you do not control, outbound to customers. Risk only if you own both legs, and then boundary rules apply. |
| Cat 8 (upstream leased) + Scope 1/2 | Risk | Under operational control a leased asset you control is Scope 1/2 and must not also appear in Cat 8. Under equity share it may appear in both with proportional allocation. |
| Cat 13 (downstream leased) + Scope 1/2 | Risk | If the lessee has operational control they report Scope 1/2 and you report Cat 13. If you retain control, you report Scope 1/2 and no Cat 13. |
| Intra-group purchases | Risk | A consolidated entity’s Scope 1 is already in the group total. Carrying the purchase into Cat 1 counts it again. |
Only the last one is unambiguously an error. The two leasing rows are correct or wrong depending on your consolidation approach, which is why the approach has to be written down beside the treatment. A reviewer reading a defensible figure with no stated basis has to test it from scratch — one sentence naming the approach turns an investigation into a check.
What this checker does not decide
It checks overlap, not completeness. Each emission appearing once says nothing about whether every category that should be present is present. Whether your inventory is complete and documented as a whole is what the Carbon Accounting Maturity Assessment tests.
It does not classify activities. Where a leased asset or a joint venture belongs is a placement question decided by your consolidation approach, and the Scope Classification Checker answers it one relationship at a time. This page assumes the placements and asks whether any emission has ended up in two of them.
It does not quantify. Whether a category belongs and what it is worth are different questions. Once it belongs, the well-to-tank calculator and the T&D losses calculator do the arithmetic, with the T&D methodology behind the second.
How we keep this current
The two factor values quoted above are read live from our emission-factor corpus rather than typed into this page, so a refresh updates them here without anyone editing the article. The rules themselves come from the Scope 3 Standard and the Corporate Standard, maintained on their own pages and linked rather than restated — with the Scope 2 Guidance deciding the transmission-loss question and ISO 14064-1 flagging the same Category 3 omission independently.
The Corporate Standard is under revision and the treatment of value-chain boundaries is among the questions in play. Nothing in the overlap rules above has moved, and that is worth re-checking rather than assuming. The tool carries the date it was run.
Frequently Asked Questions
No. Your Category 1 being your supplier’s Scope 1 is the standard working as intended, not an error to correct. Value-chain accounting overlaps across entity boundaries deliberately, so that pressure to decarbonise sits with everyone who can act on the emission rather than only with whoever produced it.
Removing it does not prevent a double count — theirs is unaffected. It only puts a hole in yours.
No — it is the same fuel across two different boundaries. Scope 1 covers burning it at your site. Category 3a covers extracting, processing and transporting it before it arrived. Neither includes the other, and both are required for full coverage.
This is the single most common reason Category 3 goes missing, and it is why the omission shows up so reliably at verification: the inventory looks careful precisely where it is incomplete.
Yes. Location-based and market-based Scope 2 both use generation-only factors, which the Scope 2 Guidance says “do not include T&D losses”. Report the losses separately in Category 3 (activity c of the Scope 3 Standard). DEFRA, for example, publishes its UK grid factor and its T&D factor as two separate rows.
The only case where a separate losses line double counts is a “delivered” Scope 2 factor that already includes losses. Switch Scope 2 to the generation factor and keep the losses line.
It depends on your consolidation approach, which is what makes this the trickiest row on the page. Under operational control, a leased asset you control belongs in Scope 1 and 2 and must not also appear in Category 8. Under equity share it can legitimately appear in both, with proportional allocation.
Because the same pattern is correct under one approach and a double count under another, the approach has to be declared alongside the treatment rather than inferred from it.
Operational control decides it, and only one party reports it in Scope 1 and 2. If the lessee controls the asset, they report it in their own Scope 1 and 2 and you report Category 13. If you retain control, you report Scope 1 and 2 and there is no Category 13 entry at all.
It cannot be both, and which one it is does not follow from who holds the title.
Not if both sit inside your consolidated boundary. The selling entity’s Scope 1 is already in the group total, so carrying the purchase into Category 1 counts the same emission a second time. Eliminate intra-group transactions on consolidation.
This is the one overlap on the page that is unambiguously an error rather than a boundary choice — and in a group that trades internally it is the easiest to create, because both entries are individually correct.
Check the five interactions rather than scanning the whole inventory. Genuine overlaps do not appear at random — they cluster in the leasing categories, in transmission losses, and in intra-group trade, because those are the places where two legitimate treatments can both apply.
Everything else that looks like an overlap is almost always inter-entity, which is the design. That is why the search is short and the fix is a boundary decision rather than a recalculation.
Under-counting, in practice, because it is harder to see and better disguised. A double count is an arithmetic error somebody can find by looking at two entries. A missing Category 3 is invisible from inside the inventory — nothing in the total signals an absence, and the reasoning that produced it felt like prudence.
That is why this checker treats a category left out to avoid an imagined overlap as the more serious finding of the two.