Attribution and double counting: whose avoided emissions are they, and how do you count them once?

Short answer. An avoided-emissions number describes the solution, not the investor. How much a fund claims for its own stake is an attribution choice, and Project Frame makes no recommendation on whether to make it. Counting once is a separate question: the same tonne should not appear twice in one number, across two methods, or along a value chain. Carbbin keeps the two questions apart.

Two questions that get mixed up

Attribution asks: of the emissions this solution avoids, how much should this investor or this component claim? It is a sharing rule, and there is no consensus on it.

Double counting asks: is the same physical reduction counted more than once? That is an error, whatever sharing rule you choose.

What the standards say

Project Frame (Guidelines 2024) treats attribution as an optional adjustment factor. Adjustment factors sit outside net unit impact and volumes and "shouldn't be mathematically blended into their underlying variables"; reports show GHG impact and adjusted GHG impact as two values (Section 3, pp. 29 to 30). Frame "does not provide a recommendation as to whether or not investors should attribute impact in their reporting" (Section 3, p. 30).

• Value chain attribution shares impact between components of a solution. Frame offers no quantification guidance and warns that impact in complex systems cannot be cleanly split: "if you remove the battery, the EV does not have 50 percent less impact; it may have none at all!" (Section 3, p. 31).
• Capitalization attribution shares impact across the funders of a company. "Frame does not advocate that investors claim a share of credit in public reports" (Section 3, p. 32), and it "does not endorse attribution as a primary method" for forward-looking impact (Section 3, p. 34).
• The joint guide by WBCSD, Project Frame and the Impact Convergence Forum (2025) notes that only the capitalization approach aligns with the GHG Protocol and PCAF. The guide itself says it is not an addendum to Frame's methodology.

The WBCSD Guidance on Avoided Emissions (v2.0) keeps avoided emissions separate from a company's inventory, with no compensation or netting. Allocation along a value chain is optional; where it is done, the allocated shares add up to no more than the end-use reduction, and the unallocated figure is reported alongside. Intermediary and end-use solutions addressing the same reduction are not summed.

ISO 14067 sets the principle that the same emissions or removals are allocated only once within the studied system (5.12). One example: a generator-specific electricity factor may be claimed by only one process (6.4.9.4.1). Carbon offsetting is not allowed in the result (3.1.1.7, 6.3.4.1). The ILCD Handbook likewise keeps offsets and system-external credits out of the boundary, the inventory and the interpretation (Provision 6.6 VII, p. 106).

How Carbbin does it

• Carbbin reports the solution's avoided emissions, per company. Every case carries the line "this is the potential of the solution, not the share attributable to any investor".
• Carbbin leaves attribution to the fund. The fund decides what to claim for its stake, and the number it starts from is one it can defend.
• Carbbin never nets avoided emissions against a footprint. They are reported as their own line.
• Carbbin counts each unit once. A full-lifetime figure and a figure inside an assessment window describe the same units and are never added together. A unit sold in one year is not counted again in a later year.

Because every company is assessed on the same method, a fund can compare and add up its portfolio. Adding up the solutions' avoided emissions is not attribution. One case needs care: if a fund holds two companies in the same value chain, a component maker and the product it goes into, their figures overlap, and the fund should present them so the same tonne is not read twice.

Common mistakes

• An ownership share multiplied into the headline, with the solution's figure no longer shown.
• Public claims of "our share" of a portfolio's avoided emissions. Frame advises against it (p. 32).
• Summing a component and the product it enables as if they avoided different tonnes.
• Adding a lifetime figure and a windowed figure for the same units.
• Counting a unit's lifetime at the sale, then again year by year.
• Two incumbent scenarios that describe the same system, both subtracted, or several incumbents added up instead of weighted by what each one actually loses to the solution.
• Netting avoided emissions against the company's own footprint, or calling the result "carbon neutral".

Sources

• Project Frame. Project Frame Methodology: Evaluating Greenhouse Gas Impact for Early-Stage Investments. December 2024. https://projectframe.how/s/2024-Project-Frame-Methodology.pdf
• WBCSD, Project Frame, Impact Convergence Forum. Avoided Emissions and Climate Investing: A Guide for Investors and Businesses. June 2025.
• WBCSD. Guidance on Avoided Emissions, version 2.0. 2025.
• ISO 14067:2018. Greenhouse gases: Carbon footprint of products. Requirements and guidelines for quantification. Clauses 3.1.1.7, 5.12, 6.3.4.1 and 6.4.9.4.1.
• European Commission, JRC. ILCD Handbook: General guide for Life Cycle Assessment, Detailed guidance. First edition, March 2010. Provision 6.6 VII.