The problem
Projects financed by multilateral development banks increasingly need to demonstrate EU Taxonomy alignment, whether because a European co-financier requires it, because the sponsor is targeting EU-domiciled sustainable bond investors, or because CSRD reporting further down the value chain pulls the requirement back to the project level. Alignment is not a disclosure exercise bolted on at financial close; it is a three-part technical test, and most MDB-financed projects that fail it do so for the same predictable reasons.
Who this is for
This is written for MDB project teams, co-financiers structuring EU-linked tranches, and sponsors preparing disclosure for CSRD-obligated counterparties. It assumes familiarity with project appraisal and focuses specifically on where Taxonomy alignment breaks down in practice.
Why this matters now
CSRD reporting obligations are pulling Taxonomy alignment questions down the financing chain: a European bank or corporate counterparty now has to disclose the taxonomy-eligible and taxonomy-aligned share of its financing activities, which means the underlying MDB-financed project has to produce alignment evidence even when the project itself has no direct EU nexus. Sponsors that treat this as a late-stage reporting task rather than a design constraint consistently find the retrofit far more expensive than early alignment would have been.
The instrument: the three-part alignment test
An activity is Taxonomy-aligned only if it clears all three tests. First, it must substantially contribute to at least one of six environmental objectives (climate mitigation, adaptation, water, circular economy, pollution prevention, biodiversity). Second, it must do no significant harm (DNSH) to the other five. Third, it must meet minimum social safeguards aligned with the OECD Guidelines and UN Guiding Principles on Business and Human Rights. Each test is assessed independently; strong performance on substantial contribution does not offset a DNSH failure.
Where projects actually stall
- Technical screening criteria treated as a checklist rather than an engineering constraint. The thresholds (emissions intensity for power generation, energy performance for buildings, water efficiency for irrigation) are specific, quantitative, and binding. Projects designed before Taxonomy alignment was a requirement often cannot retrofit their way to compliance without a material redesign.
- DNSH assessed superficially. A renewable energy project can substantially contribute to climate mitigation and still fail DNSH on water, biodiversity, or circular-economy grounds if those impacts were not assessed with the same rigour as the climate case, a common gap when the climate team and the biodiversity/water team work in isolation.
- Minimum safeguards documentation gaps. OECD Guidelines and UN Guiding Principles alignment is frequently assumed rather than evidenced, which becomes a live problem the moment a financier's compliance function asks for the paper trail.
- Data that exists but isn't structured for assessment. Most MDB projects already collect the underlying environmental and social data through standard safeguards processes. The gap is translating that data into the specific format Taxonomy assessment requires, not a lack of data itself.
All three tests must pass independently. Strong performance on one does not offset a failure on another.
Interaction with other frameworks: CSRD, ISSB, TNFD
Taxonomy alignment does not sit in isolation. CSRD-obligated counterparties need the taxonomy-eligible and taxonomy-aligned percentages for their own disclosure; ISSB-aligned climate disclosures increasingly expect the same underlying transition and physical-risk data the Taxonomy's climate objectives require; and where a project has material nature-related exposure, TNFD's LEAP approach (Locate, Evaluate, Assess, Prepare) provides a structured way to generate the biodiversity DNSH evidence the Taxonomy assessment needs. Treating these as one integrated data-collection exercise, rather than four separate compliance workstreams, is what keeps the assessment cost proportionate.
What a readiness assessment should actually produce
Not a compliance memo restating the regulation, a project-specific gap analysis: which technical screening criteria apply, where the project currently sits against each threshold, what evidence exists for DNSH and minimum safeguards, and a prioritised list of what closes the remaining gaps before the financing or reporting deadline driving the requirement in the first place.
Illustrative example: a renewable energy project failing on DNSH, not substantial contribution
Consider, illustratively, a hydropower project that clearly meets the Taxonomy's climate mitigation threshold on emissions intensity. If the DNSH assessment for biodiversity was not conducted with river-basin-level ecological data, comparable to what a TNFD LEAP assessment would generate, the project can still fail alignment outright, regardless of how strong its climate case is. This is illustrative only; outcomes depend on project-specific technical screening and DNSH evidence.
Conclusion
Taxonomy alignment is won or lost on DNSH and minimum-safeguards evidence far more often than on the headline climate metric, because sponsors invest disproportionate effort in the substantial-contribution case and treat the other two tests as secondary. Projects that commission a genuine gap analysis at concept stage, covering all three tests with equal rigour, avoid the far more expensive retrofit that follows a late-stage compliance failure.