CSRD and Scope 3: Why GHG Accounting Just Became a Data Engineering Problem

The short version: The EU's Corporate Sustainability Reporting Directive (CSRD) turns greenhouse-gas disclosure into audited, financial-grade reporting. Under European Sustainability Reporting Standard ESRS E1, in-scope companies must report Scope 1, 2, and 3 emissions, including every significant Scope 3 category — and they cannot skip a material category just because the data is messy [1][2][3]. For most companies, Scope 3 is the largest and hardest part of the inventory, which makes this fundamentally a data problem.

What CSRD actually requires

CSRD requires in-scope companies to report against the ESRS. The climate standard, ESRS E1, sets the emissions disclosure rules [1][2]. In practice, that means:

• Disclose gross Scope 1, 2, and 3 emissions, including a breakdown by each significant Scope 3 category [2][3].
• Document methodology — the calculation methods, reporting boundaries, data sources, and assumptions used [2].
• Justify exclusions — list which Scope 3 categories are included and excluded, with a reason for each exclusion [2].
• Track year over year to demonstrate progress against targets [3].

Crucially, the bar is comparable to financial reporting: disclosures are auditable, and "the data was too complex" is not an accepted reason to omit a material category [3].

Scope 3 is where the difficulty — and the emissions — live

The GHG Protocol's Corporate Value Chain (Scope 3) Standard defines 15 categories of indirect, value-chain emissions, split between upstream (purchased goods and services, transport, business travel, commuting, waste) and downstream (use and end-of-life of sold products, investments) [4]. For many organizations, Scope 3 is the main component of the total GHG inventory [2].

That is exactly what makes it hard. Scope 1 and 2 come from meters and bills you control. Scope 3 depends on data from suppliers, logistics partners, and customers — sources that range from precise supplier-specific figures down to spend-based estimates [2]. Credible, auditable Scope 3 reporting requires activity-level data, not high-level averages.

Who is in scope after the "Omnibus" changes

The timeline has shifted. The first wave — large public-interest entities with 500+ employees — applies the rules from financial year 2024, with reports published in 2025 [1]. The European Commission's February 2025 "Omnibus" / "stop-the-clock" package then proposed to narrow scope toward companies with 1000+ employees and postpone later waves, with large companies reporting from around 2028 and the entry into application for the previously-scheduled waves delayed [1][3].

The exact thresholds and dates depend on the final adopted legislation, so confirm your obligation against the current text — but the direction is fewer, larger reporters and more time, not a removal of the Scope 3 requirement itself.

Why this is a data engineering project

If you treat CSRD as a once-a-year spreadsheet exercise, you will fail an audit. Defensible disclosure requires a system that:

• Ingests activity data from procurement, logistics, facilities, and supplier surveys into a consistent model.
• Applies traceable emission factors with versioning, so a number can be reproduced and explained months later.
• Distinguishes primary from estimated data and records the method per category, exactly as ESRS E1 expects [2].
• Supports year-on-year comparability and target tracking [3].

That is an ETL, data-quality, and lineage problem before it is a sustainability problem — which is why carbon accounting increasingly looks like any other regulated data pipeline.

What to do next

Map which Scope 3 categories are material for your business, audit where that activity data lives today, and build the pipeline that makes it auditable. At Datanerds we build exactly this kind of GHG/ESG accounting engine — activity-based, factor-versioned, and reporting-ready — so disclosure becomes a query, not a quarterly fire drill.