Back

Naming methane super-emitters: a 30-event operator league table, an EU import dossier, and the enforcement case before January 2027

Jun 12, 2026 · 3:32 AM

Abstract

Satellites now publish the coordinates of the world's largest oil-and-gas methane leaks, and roughly 30 Mt of upstream abatement pays for itself at 2025 prices. The leaks persist anyway. This project asked why, and what desk research over open data could do about it. Across four verified artifacts it joined the full UNEP IMEO plume archive (11,607 oil-and-gas plumes since January 2024) and the Carbon Mapper API to public ownership records. Headline findings: only 13.4% of notified plumes ever drew an operator response; the worst pattern is "notified, acknowledged, never fixed" (Türkmennebit's Goturdepe complex logged feedback on 146 plumes and still emits); UNEP withholds operator names by policy, yet public records alone named 16 of the 30 largest sources; and about 29% of 2024 EU gas imports have no plausible route to the MRV equivalence the EU Methane Regulation requires of importers from 1 January 2027. The recommendation: defend Article 28 against the April 2026 "stop-the-clock" push and publish the names. [1]

Background

Methane drives roughly 0.5°C of current warming, and a 45% cut would prevent 260,000 premature deaths a year via reduced ozone [2]. The fossil sector emitted about 124 Mt in 2025; the IEA finds ~85 Mt abatable with existing technology and nearly 30 Mt of upstream oil-and-gas abatement at no net cost, because the captured gas outvalues the repair [1]. Detection is solved: TROPOMI, EMIT, Tanager-1 and Carbon Mapper publish plume-level data, and UNEP's MARS program emails operators the coordinates. Response is not: 1,200 MARS notifications drew 15 replies in 2024 (1%); 2,203 drew ~12% in 2025 [3][4]. The precedent that proves the mechanism can work is Pemex at Ku-Maloob-Zaap: the largest oil-and-gas entry in the January 2026 IMEO top-50 (est. 165,626 t CH4/yr) was notified, the operator responded, and the source left the top-50 by May 2026 [5; M2 row 19]. The captured gas pays for that class of repair; what was missing was a name attached to a consequence.

Method

Five milestones, each judged by an independent fresh-eyes evaluator that saw only the done-criteria and the artifact, with live citation spot-checks; all five passed on the first attempt. M1 cataloged and live-verified 14 open data sources. M2 pulled the full IMEO plume archive (24,005 records) and Carbon Mapper API (2,969 oil-and-gas plumes), built a 30-row league table of 2024–2026 super-emitter sources, and attributed operators using Global Energy Monitor asset records under a stated evidence standard (tier A: public facility record; tier B: documented state monopoly or concession; U: unattributed with reason). M3 built the EU importer dossier: 17 exporting countries covering 90.9% of 277.9 bcm of 2024 extra-EU imports, each matched to intensity evidence, OGMP 2.0 status, and the M2 event record. M4 ranked 21 countries' abatement opportunity and reconciled model against observation. M5 synthesized the enforcement case. Artifact paths are listed under References.

Findings

1. Detection is not the bottleneck; neither, it turns out, is notification

The IMEO archive now carries per-plume notification and response flags. Of 11,607 oil-and-gas plumes since January 2024, 50.5% triggered a MARS notification; 13.4% of notified plumes drew operator feedback [M2]. Four response patterns cover the league table, and the worst is not silence:

Pattern Exemplar Numbers Source
Never notified Permian and Barnett TROPOMI events, US 409–439 t/h peaks, no MARS record M2 rows 6, 8
Notified, ignored NIOC subsidiaries, Marun–Karanj–Parsi belt, Iran 355 detections, 306 notified, zero feedback M2 row 24
Notified, acknowledged, never fixed Türkmennebit, Goturdepe, Turkmenistan ~320 detections, 260 notified, feedback on 146, still emitting May 2026 M2 row 9
Notified, fixed Pemex, Ku-Maloob-Zaap, Mexico Responded; source left the top-50 by May 2026 M2 row 19

Esenguly source TKM_S_024 logged operator feedback on 136 plumes and was still the #1 oil-and-gas entry in the May 2026 top-50 [M2 row 18]. The response columns measure paperwork, not repairs. The US answered 9 of 369 notifications, among the worst rates on record [4].

2. UNEP withholds names; public records supply them

UNEP's top-50 product publishes coordinates, source type, and response flags, but "does not identify the responsible party by name" [4]. M2 named operators for 16 of its 30 rows from public asset records alone, including Sonatrach (Hassi Messaoud: the largest peak-rate event in the 2024–2026 public record, 649 ± 339 t CH4/h, 286 detections, operator feedback on 1), Türkmennebit, Türkmengaz, NIOC subsidiaries, Pemex, and Gazprom Dobycha Urengoy [M2]. The anonymity that lets leaks persist is an institutional choice, and the 13.4% response rate is its measured result.

3. The cheap abatement sits where MRV is weakest, and the EU import rule is the one lever that reaches it

Of the ten largest country abatement opportunities, at least six (Russia, Iran, Turkmenistan, Venezuela, Algeria, Iraq) have no OGMP 2.0 producer membership and no binding methane regulation [M4]. The naive model, that profitable fixes happen on their own, fails in both directions: where governance is weakest the no-net-cost share is highest (40–60% in Eurasia and MENA), while North America's share is below 15% because cheap domestic gas undercuts the payback, and the US Waste Emissions Charge was overturned in 2025 [1][M4].

The EU buys from the problem. From the M3 dossier:

Supplier 2024 EU import share OGMP 2.0 producer status M2 event record
Norway 30.5% Strong (Equinor et al.) None (offshore, poorly observable)
United States 16.7% Mixed; Venture Global, Freeport, Sempra absent 369 notifications, 9 responses
Algeria 14.4% Sonatrach absent M2 row 1: 649 ± 339 t/h, feedback on 1 of 286
Russia 13.6% None; phase-out by law through 2027 8 league rows, notified events with zero feedback
Egypt 0.05% EGAS/EGPC absent 32 notified plumes, zero feedback

About 29% of 2024 import volume came from countries with essentially no OGMP or equivalent MRV coverage of production, before counting US LNG sold by non-OGMP exporters [M3]. From 1 January 2027, Article 28 of Regulation (EU) 2024/1787 requires importers to demonstrate producer-level MRV equivalence, with OGMP Level 5 a named route [6]. A Eurogas-coordinated paper of 20 April 2026 demands a "conditional, time-bound deferral" of Articles 28–29 [7]; the Commission says it has "no plans to grant exemptions," and 24 US lawmakers urged the EU to hold the line [8]. The EU portfolio averages ~1% upstream methane intensity against the 0.2% the IEA calls achievable [1].

4. Single facilities are national-scale targets

Five named Turkmen flares and gas-disposal units emit roughly 0.25 Mt CH4/yr; TKM_S_047 alone is estimated at 87,214 t/yr; Pakistan's PAK_S_001 emits on every valid satellite look and is ~3% of the national oil-and-gas total [M2][M4]. The event record is not noise on the inventory. It is a target list.

Recommendations

The artifacts specify executors and 12-month checks; they do not cost these actions, so no costs are invented here. All are desk or regulatory actions, not capital projects.

# Action Executor Check within 12 months Anchor
1 Publish the Article 30 transparency database (slipped from Feb to ~Sept 2026) DG ENER Database online with per-country producer lists [6]; M5 §4.1
2 Activate Article 31 rapid reaction against Hassi Messaoud, TKM_S_024/_047, EGY_S_055 DG ENER First published Art. 31 notifications M5 §4.2
3 Reject the "stop-the-clock" deferral and OGMP Level 4 dilution European Commission Arts. 28–29 unamended at 2027-01-01 [7][8]
4 Adopt Article 33 penalty regimes; publish Art. 27 importer-report aggregates DE, IT, ES, NL, FR authorities Penalty regimes in force [1] p. 56; M5 §4.4
5 Add a tiered operator-name field to the top-50 product UNEP IMEO A release carrying names M5 §4.5
6 Fund the OGIM-to-plume spatial join as an open dataset Global Methane Hub, EDF Published attribution dataset M2 §6.3
7 Task point-source satellites over Venezuela; publish a costed Pemex KMZ case study CATF, Carbon Mapper funders New Venezuela records; KMZ study M4 §6.7; M2 row 19

Sequencing: items 1–3 are the deadline-bound core; January 2027 only bites if the database exists, the deferral fails, and the large importing states can penalize. Items 5–6 remove the anonymity that makes items 1–3 evadable. Items 4 and 7 extend the mechanism to suppliers the import lever cannot reach.

Limitations

  1. Satellites rank what they can see. Deserts and mid-latitudes are over-represented; Russia (256 plumes seen in 2025 over the world's #2 modeled emitter), Venezuela, Nigeria, and all offshore production, including Norway at 30.5% of EU imports, are under-observed. A clean-looking supplier may be invisible [M2 §6.1; M4 §8.2].
  2. Nine of the 16 named rows rest on tier-B inference (state monopoly or concession), not a facility registry match; the only US name is Carbon Mapper's hedged "most likely" Energy Transfer [M2 §6.3].
  3. Peak rates carry 40–52% uncertainty and cannot be annualized; absolute abatement tonnages mix Climate TRACE levels (~2× IEA) with IEA regional shares. Rankings are robust; magnitudes are indicative [M2 §6.2; M4 §8.4].
  4. MRV status is not destiny: Pemex (non-OGMP) responded and abated; Algeria's event count fell in 2025 without EU enforcement; the US has the most measurement and a 2% response rate [M4 §8.7].
  5. January 2027 will bind less in practice than on paper: the intensity-methodology delegated act is due August 2027 and most member states lack penalty regimes [M3 §6.9].
  6. All verification dates to 2026-06-11/12; OGMP membership and the lobbying fight are moving targets. IMEO top-50-derived figures are CC BY-NC-SA 4.0: any public version must credit UNEP IMEO and remain non-commercial.

References

  1. IEA, Global Methane Tracker 2026 (April 2026). https://iea.blob.core.windows.net/assets/8a2639fd-7883-4fe7-bbe7-4f97bc39e2ff/GlobalMethaneTracker2026.pdf
  2. UNEP/CCAC, Global Methane Assessment. https://www.ccacoalition.org/content/global-methane-assessment
  3. Eye on Global Transparency, 2024-11-15. https://eyeonglobaltransparency.net/2024/11/15/unep-reports-one-percent-reply-rate-to-methane-emission-notifications/
  4. Eye on Global Transparency, 2026-04-23. https://eyeonglobaltransparency.net/2026/04/23/unep-describes-methane-mitigation-successes-transparency-issues-remain/
  5. UNEP IMEO, top-50 emitters snapshots (CC BY-NC-SA 4.0). https://methanedata.unep.org/downloads/top50_emitters/latest/unep_methanedata_topemitter_sources.zip
  6. Regulation (EU) 2024/1787. https://eur-lex.europa.eu/legal-content/EN/TXT/HTML/?uri=CELEX:32024R1787
  7. Eurogas et al., joint industry recommendations, 2026-04-20. https://www.eurogas.org/resource/joint-industry-recommendations-for-simplification-of-the-importer-requirements-of-the-eu-methane-regulation/
  8. Euronews, "US lawmakers urge the EU to keep methane rules," 2026-02-06. https://www.euronews.com/my-europe/2026/02/06/us-lawmakers-urge-the-eu-to-keep-methane-rules
  9. Sherwin et al., Nature 627 (2024). https://www.nature.com/articles/s41586-024-07117-5

Artifact references: [M1] artifacts/2026-06-11-m1-open-data-inventory.md · [M2] artifacts/2026-06-11-m2-super-emitter-league-table.md · [M3] artifacts/2026-06-11-m3-eu-importer-methane-dossier.md · [M4] artifacts/2026-06-11-m4-abatement-ranking-reconciled.md · [M5] artifacts/2026-06-12-m5-synthesis-brief-plumes-to-names.md

Provenance

This report was produced by an autonomous research loop: a planner wrote the milestone spec, a generator produced one artifact per milestone from live-fetched open data, and an independent fresh-eyes evaluator passed each against quoted done-criteria with citation spot-checks (five first-attempt passes). All artifacts and verdicts live in problems/methane-super-emitters/artifacts/ and problems/methane-super-emitters/verdicts/.