A thesis page, not a source summary. Two objections to the ICC Landscape Review raised by Martin Pompéry on 2026-08-17, checked against the Review’s own text at his insistence, and upheld. Both concern the eighth criterion, Operability, and both are construction faults in the scoping of the architecture rather than gaps in its cryptography or its rulebook.
Neither objection is about whether a ledger can be built. They are about whether the thing it is being built for has the shape the Review assumes.
1. Operability (H) demands what Comparability (C) forbids
The Review’s definition of H, §4.2, verbatim — the wording is the Secretariat’s own, since H is the one criterion not taken from the Panel’s principles statement:
“The system can be implemented directly by companies in their enterprise systems rather than reconstructed annually. It records emissions for transactions, aligned to financial periods and tending towards near-real-time where feasible, rather than a once-a-year estimate. It is scalable across products, facilities, and firms, supporting consistent application in complex, high-volume business environments and across value chains; and it improves recursively as data quality and coverage grow.”
Four demands: native in the enterprise system; transactional rather than annual; scalable; and recursively improvable. The second is the one that bites.
The objection ⟨Martin Pompéry, 2026-08-17; the wiki records it as his and endorses it⟩: PCF calculation is performed ex post and aggregated, and this is not a limitation of current practice but the method’s answer to a real problem. Averaging over a period removes temporary variance — a cold month, a maintenance shutdown, an unusually empty return leg — so that two figures from two reporters describe the same kind of thing and can be compared. The accounting side of the discipline, LCA and CO₂ accounting, settled this globally and in favour of aggregation. A criterion that demands transaction-level, near-real-time recording is asking for the variance to be put back in.
Where the Review takes the opposite side without noticing it is a side. It treats period-averaging throughout as a deficiency. Table 4 debits PACT with “period-average static figures rather than transaction-level records”. §5.2 credits E-Ledgers because it “records carbon at the moment of physical transfer, which means numbers are event-driven”. Nowhere does the document state why averaging exists, and nowhere does it register that its Criterion C — comparability — is what averaging is for.
The Review’s only near-answer, and why it does not answer. §5.2 adds that event-driven numbers “can be aggregated in any reporting window”. ⟨ Inference: this relocates the averaging rather than avoiding it. Whoever aggregates over a window is averaging; the ledger does not escape the operation, it moves it from the method into the read-out. And that is worse for comparability, not better: today the window is fixed by the Product Category Rule, i.e. by the rulebook, whereas “any reporting window” puts it in the hands of whoever queries the ledger. Free choice of aggregation window is a comparability problem added, not removed.⟩
⟨Inference: so the criteria set contains an unacknowledged internal conflict, and it sits in the one criterion the Secretariat wrote itself. This is a second finding of the same type as the objection already on record that A, C, D and G are near-synonyms — see the criteria page. Together they suggest the grid was assembled to describe a preferred architecture rather than derived independently and then applied to it.⟩
What a fair response has to concede. The Review’s underlying complaint is not empty: an annual study reconstructed by hand cannot serve a quarterly CBAM cycle, and that is a genuine operability failure. The correct target is therefore not “transaction-level” but the decoupling of the recording frequency from the reporting frequency: record at whatever granularity the enterprise system produces, and let the rulebook — not the querying party — fix the aggregation window at which a comparable figure is formed. ⟨ Inference: this is a formulation ICC could adopt without abandoning the ledger, which is what makes it worth offering.⟩
2. Services are absent from a framework meant to cover Scope 3
The objection ⟨Martin Pompéry, 2026-08-17⟩: PCF calculation applies to logistics as a service, and prospectively to nearly all services — which is most of Scope 3.
Checked. The word “service” occurs once in the entire Review, in the ICC boilerplate on the last page (“market-leading dispute resolution services”). There is no section, table row or sentence on logistics, freight as a purchased service, IT, or professional services. The document’s own vocabulary is physical throughout: “physical product flows”, chain of custody through “shared tanks, pipelines, terminals, and trading infrastructure”, mass balance over certified and non-certified inputs, “emissions recorded once at source, allocated to outputs by causal logic, transferred with legal title” when the product is sold.
⟨Inference: the whole architecture is specified for a physical good that moves through custody. A service has no product to which embedded emissions attach and no custody event at which title transfers, so the ledger’s two load-bearing operations — record at source, transfer with legal title — have no obvious referent. The E-liability rule “never remove emissions from a ledger except by transfer to another ledger” needs a defined transfer event; a freight movement has an invoice, not a transferred object.⟩
⟨Inference: this is not a gap the framework-development process closes by elaborating what is there — it is a case the design has not been tested against at all. And it interacts with objection 1: services are precisely where variance is highest and where averaging does the most work. A haulier’s per-tonne-kilometre emissions vary with load factor, routing, backhaul and weather; the reported figure is a fleet-and-period average by construction, and a transaction-level figure for a single shipment would be both computable and useless for comparison.⟩
Scope of this claim. The Review is explicit that it “necessarily focuses more on better developed areas and less on emerging areas” and that unresolved items are for the Panel. It never names services as one of the deferred items, however: the deferred list (§6.2) is removals, biogenic carbon, capital goods, circular flows, market instruments and offsets. The finding is the silence, not a stated exclusion.
Why this matters for SINE’s response
⟨Inference.⟩ These two objections are of a different kind from the confidentiality gap. That one identifies a space SINE can fill with something it has — verifiable carbon accounting. These two identify places where the architecture is under-specified for its own stated purpose, and SINE has no product to sell against them. That makes them the more credible part of a response: they cost nothing to concede to, they are checkable against the document in minutes, and they establish that the critique is not positioning.
The temporal point also has a constructive form, above. The service point, at present, does not — the honest version is a question put to the Panel: how does an accountable quantity attach to a service, and what is the transfer event?
Cross-references
- The Carbon Accounting Landscape Review (ICC, 2026) — the source
- The eight decision-grade criteria — H and C in full, and the objections already on record
- Ledger-based carbon accounting — the architecture these objections test
- Verifiable Carbon Accounting — H revised there in the light of this page
- PACT PCF Calculation Methodology — where PCRs and period scoping actually live
- The confidentiality gap — the other, differently-shaped finding
- Open tasks, item 1 — the response these feed
Sources
sources/whitepapers/2026-08-17-icc-carbon-accounting-landscape-review/2026-08-17-icc-carbon-accounting-landscape-review.md(§1 scope note; §2 implications; §4.2 criterion H verbatim; §5.2 Operability; §6.2 the deferred hard cases; Appendix B chain of custody — and the whole document cited for the absence of services)sources/whitepapers/2026-08-17-icc-carbon-accounting-landscape-review/tables-recovered-from-pdf.md(Table 3 row H; Table 4 PACT row)