One trusted emissions record for every stakeholder.
EmiSource AI turns routine production data into a source-resolved greenhouse-gas inventory, with stated uncertainty, a declared evidence tier for every source and a clear line from each figure back to its data and method. Operators, regulators, environmental agencies, finance ministries and verifiers work from the same evidence.
Upstream methane, 2018 – Mar 2026
95% interval 42.7 – 132.0 kt
- Facilities
- 3
- Largest source
- 70% flare — uncombusted methane
- Tier 2 or 3
- 92%
Mean of 2,000 Monte Carlo draws; bottom-up and activity-based, not a measurement. How it is made
- IPCC 2006 Guidelines and 2019 Refinement
- Paris Agreement transparency framework
- OGMP 2.0
- EU Methane Regulation
- National methane guidelines
Everything an emissions programme needs, in one place.
From the monthly return to the national inventory and the readiness assessment, each step is recorded, explained and open to review.
Source-resolved inventory
Twelve emission sources per facility, from routine production data, each with its declared evidence tier and a Monte Carlo uncertainty interval.
Learn moreEvidence behind every figure
Each figure links to the inputs, method version and engine run that produced it. Stored figures are re-run on a schedule and every change is recorded.
Learn moreRegulatory readiness
Criteria for the EU Methane Regulation, OGMP 2.0, the IPCC and Paris transparency rules and national methane guidelines, with sources and review status.
Learn moreMethane programmes
Operators record inspection plans, leak detection and repair surveys and repairs; environmental agencies follow progress against the national roadmap.
Learn morePathways to 2040
Scenario projections with uncertainty bands, an abatement cost curve and a ranking of which measurements would most reduce uncertainty.
Learn moreAI that supports review
Anomaly screening, review prioritisation and a cited assistant help specialists work faster. Every AI output is explained, and people make the decisions.
Learn moreA workspace for each stakeholder, on one evidence base.
Each organisation sees what its work requires and nothing more. Because every party reads the same figures, discussion can focus on interpretation and action.
Operator
For example, a licensed oil and gas operator
Prepare and file monthly returns, track methane programmes and see your own facilities' emissions.
Upstream regulator
For example, the Petroleum Commission
Review returns, follow national and facility trends, and prepare evidence-backed briefings.
Environmental agency
For example, the Environmental Protection Authority
Follow operators' methane programmes and leak repair, and support national inventory reporting.
Revenue and finance
For example, the Revenue Authority and the Ministry of Finance
Draw on consistent emissions evidence for fiscal and policy analysis.
Independent verifier
For example, an accredited verification body
Trace any figure to its inputs and method, and re-run it independently.
Platform developer and research
For example, NCEL, KNUST
Maintain the method, the factor sets and the standards mapping, and administer the platform.
Training programmes use the open Library and guided material. Access is granted by organisation, under data-sharing agreements.
From routine data to reviewable evidence.
1Data in
Monthly returns by form, file or API, checked against published rules before filing. Satellite observations and OPGEE runs for comparison.
2Physically based estimation
A versioned engine resolves each source and propagates parameter uncertainty; method changes are versioned and reproducible.
3Review
Rule checks and AI screening raise questions; specialists decide, with notes, in a shared record every party can see.
4Evidence out
Briefings with a verification code, evidence packs, audited exports and readiness assessments, each traceable to its source data.
Built and tested on Ghana’s upstream sector.
Developed under the Net-Zero Carbon Emission Project with sector oversight by the Petroleum Commission, on the country’s offshore production record. The platform is designed to be configured for any producing jurisdiction: its facilities, its institutions and its national guidelines.
See the platform’s transparency record- months of production returns
- 99
- emission sources resolved per facility
- 12
- of 0 scheduled re-runs reproduced the stored figure
- —
- planning horizon, five scenarios
- 2040
Built to be relied on.
A shared record is useful only if every party can trust how it was made. These commitments are part of the design.
Estimates are labelled as estimates
Bottom-up figures carry their basis, evidence tier and interval wherever they appear, and are never presented as measurements.
Assurance stays independent
The platform supplies the evidence; accredited verifiers provide assurance, so no institution verifies its own figures.
People make the decisions
AI raises questions and prioritises work. It does not produce headline figures or decide outcomes.
Support, not determination
Screening tests and readiness assessments inform stakeholders' own judgements. They are not compliance determinations.
Bring a shared, verifiable emissions record to your organisation.
Request access for your organisation, or contact the Net-Zero Carbon Emission Lab about a demonstration or configuring EmiSource AI for your jurisdiction.