iz · how to use the bench

Use it. Cite it. Correct it.

iz is open infrastructure for Turkish CBAM compliance. Three audiences. Three paths. Pick yours.

01 If you're a Turkish operator

You ship cement, steel, aluminum, or fertilizer to the EU. CBAM is phasing in (full enforcement 2034) and you're either paying the EU default value (almost certainly an overcharge) or building MRV. Here's how iz helps:

  1. Find your plant in the bench. Filter by city or company. If your plant is in the 20 audit-grade rows (ink outline on the map), you have a defensible per-facility Scope 1 number already worked out. If it's in the 39 cf-corrected rows, the formula capacity × route-EF × cf is your starting point.
  2. Cross-check the number against your own records. Compare to your most recent audited Scope 1 (or ISO 14064-1 verification). If we're within ±15%, ship the iz number to your verifier. If we're materially off, tell us why — we'll update the bench (or you can open a PR).
  3. Hire an EU-accredited verifier (DNV / TÜV Süd / Bureau Veritas / SGS Turkey desks). Hand them the iz analysis as supporting evidence. Their job becomes a 1-day desk review instead of a 2-week ground-up audit. You pay them less; they sign off faster.
  4. Submit verified data to your EU importer. They file the CBAM declaration using your verified number, not the EU default. You avoid the overcharge.

Important caveat: downloading the iz CSV does not legally substitute for an EU-accredited verifier audit. Step 3 is required by CBAM regulations. iz makes that audit cheaper and faster, not unnecessary.

02 If you're a CBAM verifier (DNV / TÜV / BV / SGS Turkey desk)

Your Turkey desk audits a handful of operators per year. Each audit is a from-scratch crawl of operator IARs, KAP filings, and on-site inspections. iz pre-builds the reference baseline:

  1. Download tr_bench_v0.csv. 59 facilities, 20 with audit-grade Scope 1, every cell sourced.
  2. For each operator you audit, the iz row gives you: nameplate capacity (operator-verified), audit-grade Scope 1 (or cf-corrected formula result), source PDF + page number, assurance tier (ISO 14064-1 / TSRS / operator), and per-plant ratio vs the EU default.
  3. Validate against your own on-site work. If iz is off, you correct on-the-ground. If iz matches, you've cut your prep time by ~80%.
  4. Cite TR-MRV-Bench v0 in your audit methodology section (CITATION.cff in the repo gives the exact citation format). The fact that the underlying methodology is public + peer-reviewable makes your audit more defensible upstream.

No commercial license needed. Apache-2.0. If you'd like a quarterly update or want to suggest methodology changes, email Ahmet — corrections are welcome and surfaced in the changelog.

03 If you're a researcher, policy person, journalist

The bench is a public dataset. Cite it like any open dataset:

Günaydın, A. B. (2026). TR-MRV-Bench: A public per-facility emissions
benchmark for Turkish CBAM-scope industry (v0.1) [Dataset].
https://github.com/abgnydn/iz · https://iz-b0n.pages.dev

Or use CITATION.cff which GitHub auto-formats into BibTeX/APA/Chicago. The full methodology is in PAPER_METHOD.md and PAPER_DISCUSSION.md in the repo.

If you find a methodology error, a wrong capacity, a missing disclosure — open a GitHub issue or PR. Every correction makes the bench better.

04 Why this is open and not a SaaS

The math is simple enough that no one should rent it. Three numbers from each operator's own published annual report. The right answer is for every TR CBAM-scope operator to have access to a defensible per-facility baseline — paid behind a paywall, the bench helps no one.

If CBAM enforcement at €85/tCO₂ proceeds as the EU plans, ~€2 billion/year in overcharged CBAM payments to the EU treasury can be avoided by Turkish exporters using real per-facility data. That money stays in Turkey, supports the lira's current account, lowers cost-to-produce, and makes TR exports more EU-competitive at the margin. The point of iz is to make sure no operator pays more than they should because they couldn't afford the verifier audit.

This is a one-person contribution. It's maintained by Ahmet Barış Günaydın as a public good. No fundraising, no SaaS, no NDA, no upsell. If you find it useful, the only ask is to cite it and contribute corrections back.

Questions, corrections, or just to say it's useful

Email is the front door. Methodology disagreements are especially welcome — the limitations page is long and explicit, but I'm sure I missed things.

hi@barisgunaydin.com Open a GitHub issue →