The rulebook changed on 24 July 2026
Commission Regulation (EU) 2024/1834 now governs industrial fans placed on the EU market. It replaced Regulation (EU) No 327/2011, raised the minimum efficiency grades, and — more importantly for replacement projects — it judges the complete fan as sold: impeller, motor, housing and drive electronics together, at the best efficiency point.
Most cooling-fan replacements used to be a purchasing exercise: match the frame, the voltage and the free-air number, then argue about price. For equipment sold into Europe, that shortcut closed on 24 July 2026. A replacement fan that keeps an older product line alive now has to survive two different tests. It has to cool the equipment, which was always true. And the complete fan has to clear an efficiency floor that the old Regulation 327/2011 let it miss by a wide margin — for axial fans, the minimum efficiency grade rose from N 40 to N 50 in the static measurement categories, which is ten full points of efficiency at the best efficiency point.
The regulation is deliberately system-based. The European Commission estimates that fans inside the scope — electrical input power between 125 W and 500 kW at the best efficiency point — are only about 5% of the fans in use, yet account for roughly 80% of all fan electricity consumption. That is why the rules reach inside other products: a fan built into an air handling unit, a chiller or a cooling tower is in scope even when it is never sold on its own. If your equipment ships to the EU with a fan in this power band, a fan substitution is no longer paperwork-neutral. It changes the product's conformity file.
This guide is the working checklist our engineers would run before approving a replacement fan sample for an EU-bound product: how to check scope, where the new efficiency bar actually sits, what to measure on samples instead of trusting a datasheet, and which documents to collect while the supplier is still motivated to send them.
| Change | What it means in practice | Date |
|---|---|---|
| Scope | Fans with 125 W to 500 kW input at BEP, including fans integrated into other products | Applies 24 July 2026 |
| Complete-fan assessment | Efficiency is judged on impeller + motor + housing + drive as one unit, not on parts | Applies 24 July 2026 |
| Higher efficiency grades | Minimum grade N rises by roughly 6 to 10 points depending on category | Applies 24 July 2026 |
| Embedded-fan transition | Fans built into other units get extra time for the tightened limits and extended documentation | Until 24 July 2027 |
| Part-load data | Performance curves at three speeds required for most fan types | From 24 July 2027 |
| Spare parts and repair | Spares available at least 10 years after the last unit is sold; delivery within six weeks unless otherwise agreed | Applies 24 July 2026 |
Step 0: Check whether your fan is in scope at all
Do not start with efficiency tables. Start with one number: the electrical input power of the complete fan at its best efficiency point, drive electronics included. Between 125 W and 500 kW, the fan is in scope whether it is sold loose or buried inside your equipment. Below 125 W, Regulation 2024/1834 does not apply to the fan directly — which covers most small DC equipment-cooling fans — but the host product may carry its own ecodesign or energy rules, and the customer specification may still demand the data. Above 500 kW the fan is out of this regulation's scope as well.
Two traps catch teams at this step. The first is rating the motor instead of the fan: scope is set by the complete fan's input at its best efficiency point, not by the motor nameplate. The second is forgetting the drive. An EC motor's internal controller counts as a variable-speed drive under the regulation, and its losses are part of the complete fan's input power. A fan that looks comfortably below 125 W on a motor datasheet can land inside the scope once the real drive is measured — and a fan near the top of a category can shift its efficiency grade for the same reason.
Who carries the obligation also matters before any hardware decision. The party that places the complete product on the EU market under its own name owns the conformity assessment and the CE declaration. If you integrate a fan into your unit and sell the unit as yours, the fan's efficiency evidence belongs in your technical file even though you did not manufacture the fan. That is exactly why sample validation and documentation collection — the rest of this checklist — cannot be delegated to a supplier's marketing PDF.
The efficiency bar, by fan category
The regulation sets a minimum efficiency grade N for each fan category, and the required efficiency is then computed from N and the fan's input power with a formula given in the regulation. Raising N by one point raises the required efficiency by one point at any power, which makes the 2015-to-2026 comparison easy to read: axial fans measured on static pressure (categories A and C) jumped from N 40 to N 50, and on total pressure (categories B and D) from N 58 to N 64. Backward-curved centrifugal fans with a housing moved from N 61 to N 64 (static) and N 64 to N 67 (total). Mixed-flow fans on total pressure went from N 62 to N 67. Jet fans, previously unregulated, enter at N 50.
Efficiency is judged at the best efficiency point, measured on standardised airways to ISO 5801, with the efficiency determination following EN 17166 for this power range. Two consequences follow for replacement work. First, the number that matters is not the efficiency at your operating point and not a peak copied from a brochure: it is the complete fan's measured efficiency at its own best point, in the configuration as sold. A guard, a diffuser or a drive that ships with the product belongs in the test. Second, a fan that passed under the 2015 grades can be several points short today. A legacy axial fan sitting at N 44 was legal for a decade; the same hardware is now six points below the floor, and no paperwork exercise closes that gap — only different hardware does.
| Fan category | Measurement | N until July 2026 | N from 24 July 2026 |
|---|---|---|---|
| Axial fan | A, C — static pressure | 40 | 50 |
| Axial fan | B, D — total pressure | 58 | 64 |
| Centrifugal, backward-curved, with housing | A, C — static pressure | 61 | 64 |
| Centrifugal, backward-curved, with housing | B, D — total pressure | 64 | 67 |
| Mixed-flow fan | B, D — total pressure | 62 | 67 |
| Jet fan | E | Not regulated | 50 |
A compliant datasheet is not a validated fan
Nothing in the regulation forces a technology choice. It never mentions EC as a requirement: an EC motor's built-in controller is recognised as a variable-speed drive, but compliance is decided by the measured efficiency of the complete fan. A well-designed AC fan on a properly matched drive can pass the new grades, and a poorly matched EC fan — wrong impeller for the motor, lossy controller, tested without its guard — can fail them. Teams that treat 'switch to EC' as the whole compliance strategy discover this on the test rig, late.
The datasheet problem is subtler. Catalogue efficiency is usually measured on a bare fan, at the voltage and configuration that flatters it, on the supplier's own rig. Your product adds inlet obstructions, a guard, a filter, a cable loom and a control signal that hunts. The regulation's complete-fan framing is, in effect, the regulator agreeing with what validation engineers already knew: the only efficiency that counts is the one the shipped configuration produces. Sample validation is where catalogue claims get re-measured under your conditions — at the best efficiency point for the legal check, and at your installed duty point for the thermal check, because a fan can pass one and fail the other.
There is also a clock on the data itself. From 24 July 2027, most fan types must publish performance curves at three different speeds, not just the single best-point line. A supplier who cannot produce part-load data for a fan you qualify today is telling you something about that fan's future documentation — and about your spare-parts position in year eight of a ten-year obligation.
What the sample test has to reproduce
A meaningful sample test reproduces three things faithfully. First, the complete fan as it will be sold and installed: same impeller, motor, drive, guard and diffuser, powered the way the product powers it. Testing a bare impeller-motor unit and fitting the guard later invalidates the comparison, because the guard is part of the pressure loss the fan works against. Second, the measurement chain: airflow and pressure from a standardised airway method such as ISO 5801, and true electrical input power measured at the drive input, electronics losses included. Third, the product's own duty point: the sample goes into the actual unit, at the worst realistic filter state, and the team records airflow, pressure, input power, component temperatures and noise together.
Sample count matters as much as the rig. One golden sample proves one unit. Production spread is where replacement projects quietly fail: three samples that all clear the minimum grade with margin, and agree with each other inside a narrow band, are evidence. One hero unit and two untested siblings are a hope. The checklist below therefore asks for a minimum of three samples on any check that decides approval.
The sample validation checklist
Run the eight checks in order. The first two decide legality, the next four decide whether the fan actually works in your product, and the last two decide whether it keeps working across a production run. Record the measured values, not just pass or fail — the file you build here is the same file your conformity assessment will ask for later.
| Check | How to run it | Pass condition |
|---|---|---|
| 1. Identity and data pack | Collect the supplier's declared fan category, measurement category, efficiency at BEP, input power Pe, flow qv, pressure Δp, efficiency grade N and VSD status, in writing | Every data point is present, and the rating plate, datasheet and test report agree with each other |
| 2. Best-point efficiency | Test the complete fan as sold on an ISO 5801 airway; compute efficiency at the measured best efficiency point | Measured efficiency meets or exceeds the minimum computed from the category grade N at the measured input power |
| 3. Installed duty point | Run the sample in the real unit at the worst-case filter and grille state; record flow and pressure at the system resistance | The unit's required airflow is delivered with the agreed margin (we suggest at least 5%) at the real duty point |
| 4. True input power | Measure electrical input at the drive input at the duty point, electronics losses included | Input is inside the product's power budget and within the agreed tolerance of the declared value (we suggest +5%) |
| 5. Control behaviour | Sweep the control signal (0–10 V, PWM or bus, as applicable) across the full range; test start, stop and minimum stable speed | No stall, hunting or failed starts anywhere in the range; speed feedback or alarm output works as documented |
| 6. Part-load behaviour | Measure flow, pressure and input at three speeds spanning the working range | Supplier part-load curves match the samples within the agreed band — this data becomes mandatory for most types from July 2027 |
| 7. Sample-to-sample spread | Repeat checks 2 and 4 on at least three production-representative samples | Efficiency spread across samples is within the agreed band (we suggest 2 points) and no sample falls below the category minimum |
| 8. Fit, noise and features | Fit the sample to the production mounting; measure noise at the duty point under the agreed condition; verify connector, protection rating and alarm features as documented | Fits without rework, noise is at or below the equipment target, and every documented feature is present and functional |
Paperwork is part of the product now
The 2024 regulation treats documentation as a deliverable, not an afterthought. The fan's rating data — type, efficiency, input power, flow, pressure difference and efficiency grade — has to appear on the rating plate and in the technical documentation, and fans with integrated or required variable-speed drives must be identified as such. When you qualify a replacement fan, you are also qualifying the paper trail your own conformity file will cite. Collect it during the project, while the supplier still answers emails quickly.
Spare parts deserve the same discipline. The regulation requires spare parts — impellers, motors below 10 kW, bearings, drives, stators — to remain available for at least ten years after the last unit of a series is sold, deliverable within six weeks unless a contract says otherwise, and repairs must be possible with standard tools, with instructions available to professional technicians. For a replacement project this cuts both ways: it protects you on the new fan, and it explains why the old fan disappeared. Ask the new supplier, in writing, how the ten-year commitment is covered for the exact model you are qualifying — a discontinued-in-two-years fan recreates your original problem on a schedule.
- Rating plate and datasheet carry the same category, efficiency, Pe, qv, Δp and N values — no contradictions between documents.
- The test report names the standard used (ISO 5801 airway, EN 17166 determination) and the exact configuration tested, drive included.
- Variable-speed status is explicit: integrated drive, drive required, or no drive — and the documentation matches the hardware.
- Spare-parts availability (10 years) and delivery time (six weeks default) are confirmed for the qualified model in writing.
- Part-load data at three speeds is already available, or the supplier commits to a date before the July 2027 requirement lands.
Where replacement projects usually fail validation
The failures repeat across projects, and most of them are configuration mismatches rather than bad fans. The supplier tested the fan without the drive that ships with it, so the measured input power was flattering and the complete fan misses its grade on your rig. The guard or diffuser fitted in the product was absent on the test, moving both the best efficiency point and the duty point. The sample was a hand-checked golden unit, and the third production sample lands two efficiency points lower — under the floor. The datasheet best point was recorded at a different voltage or frequency than your product supplies. Or the team validated efficiency carefully and never checked control behaviour, so the fan stalls at the low end of its speed range inside a quiet product that spends its life there.
Every one of these is cheap to catch at sample stage and expensive after tooling, printing and certification have locked around the choice. That asymmetry is the whole argument for running the checklist before the purchase order, not after the first production batch.
How MEGA Tech supports an ErP replacement review
MEGA Tech has manufactured cooling fans since 2008, with an office in Guangzhou and a factory in Shenzhen, certified to ISO 9001 with CE and RoHS documentation for its range. The support we offer for a replacement programme follows the same five steps as our custom work: custom cooling fan development and manufacturing, fan selection, airflow-path review, sample validation and volume production. For an ErP-driven replacement that means, concretely: capturing the installed duty point of the fan you are replacing, shortlisting candidates whose complete-fan data clears the new grade with margin, and running the eight checks above on samples in your equipment or in a bench setup that reproduces it.
Two boundaries are worth stating plainly. Our own catalogue is DC axial cooling fans; where a project needs a large EC plug fan or an AHU assembly from a specialist maker, the checklist still structures your sample testing of those units, but we will tell you the unit is outside our range rather than stretch a product to fit. And the conformity declaration for your product stays yours: whoever places the complete product on the EU market under their own name signs it. What we supply is the fan hardware, the measured sample evidence and the documentation pack your technical file needs — which is the part that usually goes missing.
If you are working against the July 2027 embedded-fan deadline, send the current fan's rating plate photo, the equipment's airflow or temperature problem, and the space the fan has to fit. That is enough to start the scope check and the duty-point capture.
Engineer FAQ
Q: Does ErP 2026 force us to switch to EC fans? A: No. The regulation sets efficiency outcomes for the complete fan and recognises an EC motor's internal controller as a variable-speed drive, but it mandates no technology. A strong AC design can pass the new grades and a weak EC design can fail them; the measured best-point efficiency of the complete fan decides.
Q: Our fan draws 90 W. Are we out of scope? A: For the fan regulation itself, yes — Regulation (EU) 2024/1834 starts at 125 W input at the best efficiency point. Check the host product's own ecodesign rules and your customer specification before concluding nothing applies, and note that scope is judged on the complete fan including drive losses, not the motor plate.
Q: The fan is built into our unit. Who is responsible for compliance? A: Whoever places the complete product on the EU market under their own name owns the conformity assessment and declaration. Fans integrated into other products also get a transition: the tightened limits and extended documentation apply to embedded fans from 24 July 2027.
Q: Can we rely on the supplier's test report instead of testing samples? A: Only if the report covers the complete fan in the configuration you will buy — drive, guard and diffuser included — measured to ISO 5801 with the efficiency determined per EN 17166. Even then, sample testing is what catches production spread; reports describe the unit that was tested, not the units that ship.
Q: What changes on 24 July 2027? A: Two things: the embedded-fan transition ends, so built-in fans must meet the tightened limits and extended documentation; and most fan types must publish part-load performance curves at three speeds. Qualifying a fan in 2026 without its part-load data means re-opening the file within a year.
Questions for a Design Review
Bring these answers, with measured values attached, before the replacement fan is released to production.
- What is the complete fan's measured input power at its best efficiency point, and does that put it inside the 125 W to 500 kW scope?
- Which category and measurement category apply, and what minimum efficiency does the grade-N formula produce at that power?
- By how many points does the worst of at least three samples clear that minimum — and what is the sample-to-sample spread?
- Was the tested configuration exactly the shipped configuration: drive, guard, diffuser, supply voltage and control signal?
- At the installed duty point (worst-case filter state), what flow, pressure, input power and noise did the sample produce in the real unit?
- Are rating plate, datasheet, test report and VSD status consistent, and is the ten-year spare-parts commitment confirmed in writing?
- Is part-load data at three speeds already in hand for the July 2027 requirement?
