Compared with the company-reported 7,869.19 tCO₂e — a 2.90% difference. This comparison does not show that the reported figure is wrong.
← Back to selected work
Independent case study / 2023–2025
Reconstructing a corporate GHG inventory from public data.
A Scope 1 and Scope 2 analysis of Unilever Indonesia, built to make calculations, assumptions, and evidence gaps visible.
PROJECT TYPEIndependent analysis
METHODExcel-based reconstruction
BOUNDARYOperational control
PERIOD2023–2025
This is independent portfolio work using public disclosures. It was not commissioned by Unilever Indonesia and is not an audit or assurance engagement.
01The question
What can the public record actually support?
Corporate emissions figures are useful only when their activity data, factors, accounting basis, and boundaries are understood. I reconstructed Unilever Indonesia’s disclosed Scope 1 and 2 inventory for 2023–2025 to compare model outputs with reported figures and identify where assumptions or missing evidence affect interpretation.
The work separates company-reported data, independent estimates, adopted values, and conditional results. It does not attempt to establish that the company’s reported figures are incorrect.
02Key results
A useful comparison, with important conditions.
Uses a disclosed 0.80 tCO₂/MWh JAMALI factor based on 2019 data. It is a historical proxy, not a company-published location-based value or a 2025 grid average.
*A zero result depends on I-REC quality, retirement, vintage, and load matching evidence that is not established by the public materials reviewed.
i
The workbook’s QA sheet records eight PASS, one CONDITIONAL, one REVIEW, and one NOT AVAILABLE. These are internal model checks, not external assurance.
03Approach
A model designed to be followed, not just read.
01
Trace the source data
Mapped publicly disclosed activity data and company-reported values to the relevant inventory lines.
02
Show every assumption
Separated emission factors, fuel-mix assumptions, adopted values, and independent calculations across dedicated workbook sheets.
03
Test the interpretation
Used sensitivity analysis and QA/reconciliation to examine where results depend on inputs or evidence not disclosed publicly.
04What remains uncertain
The boundaries matter as much as the numbers.
01 / ELECTRICITY
Location-based is an estimate.
The grid factor is a historical proxy. The public materials reviewed do not provide a company-published location-based figure for comparison.
02 / CERTIFICATES
Market-based zero is conditional.
A public zero-emissions claim needs supporting evidence for the instruments used, including retirement, vintage, and matching.
03 / COMPARABILITY
The boundary changed.
The 2025 inventory includes 11 warehouses versus three in 2024, and the ice cream factory for January–November 2025. Direct year-to-year comparison needs care.
04 / COVERAGE
Scope 3 was not calculated.
This project reconstructs Scope 1 and 2 only. The company’s reported 94.34% reduction KPI is noted as a company claim; this workbook does not independently recalculate it.
05Explore the work
See the evidence behind the case study.
The workbook holds the source data, calculations, assumptions, sensitivity analysis, and QA record. The executive deck gives a shorter visual walkthrough.
Source basis: Unilever Indonesia’s public sustainability disclosures and the sources documented in the workbook. Figures shown here reflect the workbook reviewed on 28 September 2026.