The boundary was never written down
Sites, entities and joint arrangements were included by habit. Without a stated consolidation approach and an entity list, completeness cannot be tested at all.
ISO 14064-1:2018
- Home
- Insights
- GHG assurance
- Six evidence gaps that delay verification
Six evidence gaps that delay verification
First-cycle GHG verifications rarely stall on arithmetic. They stall because a figure cannot be traced back to a primary record, because a boundary was never written down, or because a factor cannot be attributed to a version of a source. These are the six gaps that most often extend an engagement, and the evidence that closes each.
First published01 September 2026
Last technical review01 September 2026
Review cycleAnnually
AuthorTechnical Directorate, Energy Up International
Technical reviewCompliance and Accreditation Officer
Reading time10 min
The pattern
Where engagements actually stop.
Emission factors have no version
A factor is quoted without its source, publication year or applicable region. Two departments then use two different values for the same fuel.
Traceability
The number cannot be traced back
The reported figure exists in a spreadsheet, but the path from meter reading or invoice to that cell is broken by a manual step nobody documented.
Data flow
Estimates are not identified as estimates
A missing month was filled by extrapolation and the result was reported as measured data. The estimate may be reasonable; presenting it as measurement is not.
Uncertainty
The base year has no recalculation record
An acquisition, disposal or methodology change occurred and the base year was never revisited, so the trend being reported is not a like-for-like comparison.
Base year
Site evidence does not match the report
Meters listed in the calculation do not exist, or exist under a different tag. Nearly always a records problem rather than a data problem — but it is found on site, late.
Site work
Closing the gap
What evidence resolves each one.
| Gap | What closes it | Prepare it by |
|---|---|---|
| Undefined boundary | A written boundary statement: consolidation approach (control or equity share), entity list with ownership percentages, site register, and stated exclusions with reasons. | Before the data year, not after |
| Unversioned factors | A factor register: value, unit, source document, publication year, region of applicability, date applied, and the person who approved it. | At the point each factor is first used |
| Broken traceability | One worked example per source, from primary record to reported figure, with every transformation visible — including manual steps. | Ahead of the engagement, as a self-test |
| Unlabelled estimates | An estimation log: which period, which source, which method, why measurement was unavailable, and the estimated effect on the total. | As the estimate is made |
| Base-year drift | A base-year statement plus a recalculation policy, and a record of every event that triggered — or was assessed and did not trigger — a recalculation. | At each structural change |
| Site mismatch | An asset and meter register that is reconciled to the calculation model before the site visit, not during it. | Two to four weeks before site work |
Every item in the middle column is a document an organisation can produce on its own. None of them requires a consultant, and none of them can be produced by the verifier — a verification body that assembles this evidence for you has disqualified itself from examining it.
The underlying cause
Most gaps are governance, not measurement.
Read the six together and a pattern appears: in each case the organisation knew the answer, but the answer had not been written down at the moment it was decided. Boundary, factor, estimate, recalculation — all four are decisions, and a decision that leaves no record cannot be evidenced a year later by anyone, however competent.
The practical remedy is unglamorous. Assign one owner for the inventory who is accountable for the decision record, not only for the number. Require that every judgement is captured when it is made, with its reason. That single control removes more verification friction than any software.
It also changes the shape of the engagement. Where the decision record exists, the verifier spends the time on testing rather than on reconstruction — which is what the fee was meant to buy.
Describe the claim — not the service.
Tell us what must be relied on, who will rely on it and by what date. We will confirm the correct programme, evidence requirement and responsible entity — or tell you it is outside our scope.
Primary sources
- ISO 14064-1:2018 — organizational quantification and reporting requirements — www.iso.org/certification.html (checked 01 September 2026)
- ISO 14064-3:2019 — verification and validation of GHG statements — www.iso.org/certification.html (checked 01 September 2026)
- ISO 14066:2023 — competence requirements for validation and verification teams — www.iso.org/certification.html (checked 01 September 2026)
Editorial note. Author and technical review above are editorial functions. They are separate from the independent technical review and the decision functions required under ISO/IEC 17029:2019, which apply to validation and verification engagements and not to published articles. Nothing on this page is advice on a specific engagement, and nothing here extends or replaces the accreditation records.
Read next
Describe the claim — not the service.
Tell us what must be relied on, who will rely on it and by what date. We will confirm the correct programme, evidence requirement and responsible entity — or tell you it is outside our scope.