EN 17038-3:2026: The Efficiency Index Moves from the Pump to the Package

A pump's efficiency used to be a property of the pump. The European standards family for pump energy performance has now completed the next step: in September 2026, CEN published EN 17038-3:2026, Methods of qualification of the Energy Efficiency Index for rotordynamic pump units — Part 3: Testing and calculation of energy efficiency index (EEI) of booster sets. The document defines how a packaged assembly of pumps — the booster set — earns a single efficiency number. This article records what the part covers, why the shift from pump to package matters beyond its building-services scope, and where the public record stops. It is not a compliance determination.

What is a booster set, in the standard's own scope?

A pump assembly sold and commissioned as one product, inside a building, moving clean water.

The catalogue record draws the boundary precisely: a booster set is a single pump unit or an assembly of pump units in parallel, up to 150 kW of hydraulic power, with a rated flow of at least 6 m³/h, with backflow prevention and the components needed to control pressure or provide flow in open loops inside buildings — placed on the market as one product, pumping clean water, without self-priming function. Sets below the 6 m³/h line are handled under Part 2 of the family.

That scope is building services, not process industry — and saying so plainly is part of reading the document correctly.

How is the index actually determined?

Not from a single operating point — from a reference load pattern, reached by either of two documented paths.

The method anchors on the 100%-duty point — the duty point of maximum hydraulic power at nominal speed — and from it derives a reference pressure-control curve and a reference flow-time profile: how the set is expected to be loaded across representative operation. The average electric power input over that profile is then established in one of two ways: by physical testing on a bench worked to EN ISO 9906 discipline, or by a semi-analytical model (SAM) that builds the same answer from the pumps' documented behaviour under fixed and variable speed. Four annexes carry the supporting material — configurations and control modes, the effect of control type on the index, the effect of control deviations, and the uncertainties and tolerances of the result.

The consequence is structural: two booster sets can be compared on one number, and the number means the same thing for both.

Why does the direction matter beyond building services?

Because the rated object has changed — and packaged units are where industrial pumping is going too.

A pump's catalogue efficiency is measured on the pump alone, at its best point, on a test stand. A booster set's EEI is measured on the assembly as it will actually run: the pumps, the controls, the valves and the pipework between them, across a realistic load pattern. The two numbers answer different questions, and the second one is the one the electricity bill asks.

The process industries are not governed by this document — its scope is explicit. But the direction it formalises is the same direction industrial procurement has been drifting for years: skids and packages rather than bare machines, and performance guaranteed on the package rather than the component. A standard that teaches a market to read a package-level efficiency number trains buyers everywhere to ask for it.

Where does the public record stop?

At three places.

The method's substance is in the paid text. The control curves, the profile, the SAM's construction and the uncertainty treatment live in the document itself; the free record carries its scope and structure.

The scope is not process pumping. Nothing in EN 17038-3:2026 applies to chemical-process duties; drawing an analogy is analysis, not citation.

An index is not a lifecycle cost. The EEI enables comparison of booster sets on one number; it does not price energy, maintenance or failure risk over a plant's life.

What does the trend add up to?

The European pump-efficiency framework keeps widening its object: from the bare pump (EN 17038-2), to the pump assembly as a marketed product (Part 3). And the ecodesign review running alongside it — the rebuilding of the water-pump rules this site covered earlier — now has the package-level measurement method in place before the package-level requirement can be written.

The durable lesson is older than the standard: efficiency claims are only as good as the boundary drawn around the machine they are measured on. EN 17038-3:2026 redraws that boundary around the package.

What this article does not cover

Note that this article does not reproduce the standard's methods, does not compute or compare any product's EEI, does not evaluate any pump or booster set, and gives no prices. It records the publication and the public catalogue record of a new measurement standard. It is not a compliance determination.

Where the next constraint sits

The regulatory track this measurement standard feeds into — the review of the EU's water-pump ecodesign rules and what it changes — is covered in Europe Is Rebuilding Its Water-Pump Ecodesign Rule: What the 547/2012 Review Changes.

And before efficiency is argued about at all, the duty has to be read in the right order — risk first, medium second, duty point third — as set out in Chemical Pump Selection Guide: Risk, Medium and Duty Point — in That Order.

Sources