What six reports reveal about emission factor provenance
A report can name an emission-factor publisher without identifying the exact factor used in a particular cell. It can also identify a version while leaving the relationship between two reporting years unclear. This study examines six selected reports to ask how far their own disclosures carry a reader toward a reproducible factor record.
The sample is purposive and historical. Each report supplies one case, and every case is limited to the cited table and notes. This is neither a survey of disclosure quality nor a recommendation to adopt any reported factor. We have not tested the companies’ activity data, reproduced their inventories or inspected the underlying factor libraries on their behalf.
The original cohort came from a larger local evidence edition than the separately published 1,040-record projection; the selected Mrs. Bectors records were outside that projection. This fresh study preserves those six cases and checks the linked reports directly. It does not claim that every underlying record was in the public dataset.
Six cases in a provenance matrix
Mazagon Dock: references attached to different scopes
Mazagon Dock’s 2024–2025 BRSR names India GHG Program road-transport factors for company-owned vehicles and CEA version 20.0, December 2024, for Scope 2. The references have different jobs; one factor-source label should not be applied across the table.
KRBL: a current-year marker
KRBL’s 2024–2025 BRSR links its marked current-year Scope 2 cell to CEA version 20.0. The unmarked comparative does not establish the same factor assignment; keep that year’s provenance unresolved.
Gateway: a source name without an identified edition
Gateway’s 2024–2025 BRSR describes a CEA average grid factor including renewables without specifying its edition or value. Asterisks sit on year headers, while notes discuss emission scopes. Preserve this ambiguity; the markers do not supply a reliable factor-to-year mapping.
Vardhman: a revision is part of the history
Vardhman’s 2024–2025 BRSR describes revised prior-year calculations using a CEA version 19.0 factor including renewables and captive-power injection. It also describes a grid-factor change from 0.716 to 0.727 tonnes CO2e/MWh. The current factor’s database edition is not identified in the examined note. Retain that limit and the revision history; neither factor is independently validated here.
eClerx: numeric factors linked to reporting years
eClerx’s 2024–2025 BRSR maps CEA version 20 and 0.842 kg CO2/kWh to 2024–2025, and version 19 and 0.823 to 2023–2024, for India. Preserve CO2 rather than silently substituting CO2e.
The factor difference is 0.019 kg CO2/kWh, or 2.31% of the earlier factor. This original sensitivity holds every other input fixed; it cannot explain the issuer’s actual emissions change.
Mrs. Bectors: version and restatement context together
Mrs. Bectors’ 2025–2026 BRSR identifies CEA version 21 for Scope 2, manufacturing-location coverage and restated 2024–2025 figures. Read the factor reference with that coverage and revision context. Do not extend this note to older report editions or infer how much emissions movement it explains.
What the matrix can establish
The cases address three separate reading questions: which calculation a reference belongs to, how precisely the source is identified, and which reporting year or revision it concerns. Keep these as separate fields. A longer citation may make a factor easier to locate without establishing that it was suitable, applied correctly or used for every activity in the reported total.
Code only what the examined passage establishes. “Publisher named”, “edition named” and “numeric factor stated” describe different evidence. If a field is unresolved, record the missing relationship rather than assigning a poor-quality score. A table-level note can support a narrower conclusion than a company-wide methodology statement. Absence from the selected passage is not proof of absence everywhere.
Follow a six-link evidence chain
1. Start with the reported cell. Identify the metric, scope, reporting year, geography and population. Keep its exact unit and table locator. This prevents a factor explanation for one row from being attached to a different row merely because both appear on the same page.
2. Connect the cell to its issuer explanation. Record the footnote marker, relevant sentence or method reference. Preserve qualifications and the source edition. A nearby note is not automatically applicable to all columns; identify the relationship that the report actually states, and leave uncertain links open.
3. Identify the named source. Keep publisher and dataset or document title separately. An organisation can publish several factor sets for different purposes. A publisher name alone is therefore a starting point for retrieval, not a unique factor identifier or a substitute for a selection rationale.
4. Preserve the edition and variant. Record the release, correction state and applicable data vintage when available. Distinguish the issuer reporting year from the factor’s data year and publication date. Do not replace the historical factor record with today’s version simply because the current website presents it first.
5. Locate the factor within that edition. Retain the row or series, original precision, units, gas basis and represented activity or geography. If the report does not provide enough information to identify the factor, state that limitation. Reconstructing a plausible number from the total is a separate inference, not a recovered source citation.
6. Bind the factor to the calculation. Identify the relevant activity inputs, conversions and calculation version. A complete chain lets another reviewer locate the intended inputs and inspect the relationship. It does not establish their accuracy or approve their use. Keep any later replacement and its reason linked to the earlier record.
Original decimal-precision sensitivity test
All quantities and factors in this exercise are invented and must not be used for real reporting. A fictional calculation uses 1,250,000 kWh and a factor recorded at 0.4176 kg CO2e/kWh. Assume the activity, boundary, period and units are matched. The only change being tested is replacing the recorded factor with a rounded display value.
Using the recorded precision gives 1,250,000 × 0.4176 = 522,000 kg CO2e, or 522 tonnes CO2e. If an export rounds the factor to three decimal places and the next calculation uses 0.418 instead, the result becomes 522,500 kg CO2e, or 522.5 tonnes. The difference is 0.5 tonne, with no change in activity.
Both multiplications are correct for the numbers entered. A result-only arithmetic check cannot decide which factor representation was intended. The source value and calculation input must remain distinguishable from a shortened display. Preserve the precision actually supplied; adding extra decimal places would not create better evidence.
Repeat the test by converting the activity to 1,250 MWh and the original factor to 417.6 kg CO2e/MWh. The result remains 522,000 kg. This equivalent-unit check can detect an inconsistent conversion, but it cannot repair a missing edition, uncertain applicability or mistaken cell-to-footnote link. Arithmetic consistency and source traceability answer different questions.
Use the chain without overstating the result
For a research comparison, record the source fact, your interpretation and the remaining evidence request separately. Where an issuer describes a revised factor or comparative, retain both the reported explanation and its limits. Do not attribute the whole emissions movement to the factor unless separately supported inputs and a defined comparison establish that relationship.
The six-link chain is an original reading aid, not a prescribed reporting form. Use the emission-factor selection guide for the separate question of whether a candidate fits an intended calculation. This study does not select production factors, establish regulatory compliance or provide an independent assurance conclusion.
Sources checked 15 October 2026. Mazagon Dock: printed/PDF 39, notes 17–18.
KRBL: printed/PDF 34, Scope 1–2 factor notes.
Gateway: PDF 20, printed 19, emissions notes.
Vardhman: printed/PDF 32, emissions methodology notes.
eClerx: printed/PDF 25–26, emissions table and year-specific factors.
Mrs. Bectors: printed/PDF 28–29, emissions table and notes.
Selected issuer facts are attributed to their original reports, which retain their own rights and qualifications. No report assets, factor-library tables or third-party assurance material are reproduced. The reading method, fictional exercise and artwork are original Nitium material. The issuers are not presented as clients or endorsers, and this study creates no new company-data records.
Comments