What Eurovent ECP-05-2026 Requires From AHU Selection Software

There is a persistent misunderstanding about Eurovent certification for air handling units: that it is a quality mark awarded to a box. It is not. The Eurovent Certified Performance programme for AHUs, whose current technical certification rules are ECP-05-2026, certifies that the performance a manufacturer declares matches what an independent laboratory measures on a unit drawn from the range. The certificate is about the accuracy of your numbers.

Which means the certification burden does not land on the sheet metal. It lands on the software that produces the numbers — the selection tool, and the technical data sheet it prints. If that tool rounds a class boundary the wrong way, or evaluates one season instead of two, the discrepancy shows up in an audit even though the unit itself is exactly what you thought you sold.

Here is what the programme actually demands of the software layer.

1. The class scale is A+ to E — and E is the bottom

ECP-05-2026 defines energy classes from A+ down to E, evaluated through the fs_Pref ratio: the unit's actual internal specific fan power against the reference value for the class being claimed. A unit passes a class when fs_Pref ≤ 1.0 against that class's reference row.

The detail that catches software out is at the bottom of the table. Classes A+ through D each have a reference row to be checked against. Class E has none — the rules state that the lowest classes carry no requirement, and no calculation is required to reach E. E is the residual class: it is always attainable, and fs_Pref is 1 by definition.

This is a five-minute test you can run on any tool. Feed it a deliberately bad unit — a high-velocity casing with an undersized fan — and see what it prints. If it reports a class the standard does not define, the class engine was written from an assumption rather than from Table 3. We have seen this specific failure produce "class F" on units that were legitimately class E, against a customer's own Eurovent sheet.

2. Both seasons, every time

An AHU that comfortably passes in January can fail in July. Air density, heat recovery duty and the coil-side pressure drops all move with the operating point, and the internal SFP moves with them. ECP-05-2026 works from seasonal evaluation, so a tool that computes one operating point and prints one class is answering half the question.

The practical requirement: your software should evaluate and display both the winter and the summer case, and should tell you which one is governing. The winter vs summer breakdown covers where the two diverge, and by how much.

3. ErP is a separate law — do not conflate the two

Regulation (EU) 1253/2014 and its ecodesign successors are EU market law: a non-residential ventilation unit either meets the minimum requirements or it cannot be placed on the market. Eurovent certification is a voluntary programme about data accuracy. A unit can be ErP-compliant and uncertified; a unit can be certified and, on a given configuration, fail an ErP threshold.

Software that merges the two into a single green tick is hiding a distinction your customer's consultant will not miss. Both verdicts should be visible, separately, with the inputs that produced each. Our ECP-05-2026 and ErP compliance guide sets out where the two regimes overlap and where they do not.

4. The audit trail is the deliverable

Because the programme tests declared data, every published number needs a provenance. In practice, four requirements on the tool:

Requirement	What the software must be able to show
Traceable component data	Each fan, coil and filter figure attributed to the supplier engine that returned it, not to a fitted curve
Stated operating point	Airflow, external static, air temperatures and densities printed beside the result they produced
Reproducible class calculation	The internal SFP, the reference value used, and the resulting fs_Pref — not just the letter
Declared vs. absent	A number the supplier does not publish shown as absent, with a reason, never as a plausible default

That last row is the one that decides audits. A selection tool that quietly substitutes a "typical" 1.2 kg/m³ air density, a 0.9 drive efficiency or a standard 1200 × 600 coil face will produce a data sheet that looks complete and cannot be reproduced by a laboratory. If a number was never published, the honest output is an empty field with a sentence saying why. The data sheet reading guide shows what a consultant looks for.

5. Questions worth asking in a demo

• Show me a class E unit. Then show me what you print when the fan power is invalid.
• Where is the summer class, and which season governs this selection?
• Which supplier engine returned this fan efficiency — and what happens if it returns nothing?
• Print the internal SFP and the reference value, not only the letter.

AirSelect3D computes ECP-05-2026 classes for both seasons and prints the inputs behind each one — it is Eurovent-ready in the sense that it produces auditable declared data. The certification itself belongs to the manufacturer and the unit range, not to the software.

Run a compliance check on a real selection →