iz · Bench · Gebze Dilovası Plant

steel · Dilovası · 4,500,000 t/y

Gebze Dilovası Plant

Çolakoğlu

01 Snapshot

Audit-grade Scope 1 · 2024
566,519
tCO₂ · direct · operator-audited financial-grade
cf-corrected formula (leave-one-plant-out)
539,539
-5% vs. truth
EU CBAM default
8,550,000
+1409% vs. truth
95% conformal prediction interval (formula, leave-one-plant-out, n=19): 291,976997,009 tCO₂ (truth inside ✓)
95% interval restricted to steel-EAF calibration set: 521,065558,668 tCO₂ (truth outside ✗ · undercoverage caveat at small n)

Open printable audit summary → One-page PDF you can hand to a verifier.

02 Facility data

Company
Çolakoğlu
Plant
Gebze Dilovası Plant
Sector
steel
City
Dilovası
Capacity
4,500,000 t/y
Label source
disclosure
Provenance
Per-plant Scope 1 directly disclosed by operator.
Headline method
cf-corrected formula (capacity × route-EF × cf), evaluated leave-one-plant-out. +82.3% log-MAE reduction vs EU default across the 19 validatable plants.

03 Source citation

Çolakoğlu 2024 SR p82

Single-site EAF plant — direct disclosure

05 Hyperspectral coverage

2 EnMAP L2A scenes index this plant (0 with <1% cloud cover). Source: DLR EOC STAC. Following Borger et al. (2025), these scenes can yield direct CO₂ and NO₂ plume retrievals at ~30 m spatial resolution.

Lowest cloud cover scenes (showing top 5):

Downloads require a free DLR EOWEB account (see enmap_access_howto.md).

06 Verify / correct

If you're an operator or analyst with a more recent disclosure than what's shown above:

Corrections that include a verifiable public source land within 24 hours. All data is Apache-2.0 licensed; methodology at iz-mrv.pages.dev/paper.