wbd-rd.nl · R&D lab · supplier information

Control-valve data we need for independent evaluation

owner René de Ren published 2026-08-04 version 1.3 draft
We select control valves (aeration and comparable low-Δp gas duty) by recomputing their performance at our site conditions, not the conditions of your example sheet. As a publicly owned organisation every number in an evaluation must be traceable and reproducible from the supplier's own data. This page shows the dataset that makes that possible, and the check we run on it before a valve enters evaluation.

What we compute from your data

Fictional example — every number and curve on this page describes an invented DN200 valve. It is illustrative only and is not any manufacturer's product.
values as a table

An example table pins us to your chosen pressure and temperature. A resistance coefficient ζ per stroke lets us regenerate this entire figure at any condition — move the sliders above: that recalculation is exactly what we must be able to do ourselves, for every valve in a comparison, on identical terms.

Inherent versus installed, and what we derive from your curve

A bench curve is the inherent characteristic: flow against stroke at a constant pressure loss across the valve. In a plant that pressure loss is not constant. As the valve opens, flow rises and the rest of the system — pipework, diffusers, the header controller — takes a larger share of the available pressure, so the valve keeps less of it. The same valve therefore shows an installed characteristic flatter than its bench curve: a strongly equal-percentage valve can end up close to linear. The two are linked by valve authority a, the share of the flow-dependent pressure loss sitting in the valve at that duty:

the link

d(ln Q)/d(stroke) = a × d(ln Kv)/d(stroke) the right-hand factor is the valve's, the authority is the system's

That split is the reason we ask for ζ and Kv instead of a flow table, and the reason we compute installed behaviour ourselves against our own measured duty: your dataset stays site-independent, and every supplier is compared on identical terms. Send the inherent curve — the installed one is ours to calculate.

Size-free comparison metrics

From the same ζ/Kv table we derive the figures below for every submission — nothing to fill in, and nothing we ask you to declare. They are dimensionless, so a DN125 and a DN500 can be judged against each other: nominal size is a design variable, not a category. Everything is read at 90 % stroke, because a control valve is not sized to sit at full travel, and because close to 100 % most constructions stop gaining area — the last two rows say whether yours does, straight from your own numbers. You can watch them being computed in the submission form.

metricdefinitionwhat it tells us
rd(ln Kv)/d(stroke) at 90 % stroke, % per % stroke the exchange rate between dosing resolution and throttling loss
stroke for a 10 % capacity changeln(1.1) / r how much travel a 10 % flow correction costs
capacity step per 0.5 % of stroke0.5 × r what one positioner step does at the reference stroke
controllable stroke rangestrokes where 0.5 % of stroke moves capacity ≤ 5 % the travel that is actually usable for control
capacity range coveredKv ratio across that range the turndown reachable inside the controllable travel
gain uniformityrmax / rmin across that range 1 is perfectly uniform; a large value means the loop is retuned by the duty
capacity reached at 90 %Kv(90 %) / Kv(full travel) what the valve still has in reserve above the sizing point
tail gain ratiolast stroke step's ΔKv ÷ the largest step's ΔKv 1 = the valve keeps gaining area to the end; well below 1 = the final travel is geometry, not capacity, and reading anything at 100 % stroke is misleading

None of these score a valve on its own. A low r is fine resolution; a high r buys range in less travel. What decides the choice is where those numbers land at the duty we actually have, together with the pressure the valve costs there — and that calculation is ours to make.

The dataset

Tier 1 · required

1

ζ versus stroke

Resistance coefficient at ≥ 10 points from 10 % to 100 % stroke, with the reference area named (nominal bore or seat area).

2

Kv versus stroke

Kv in m³/h at the same stroke points. Deliberately redundant with ζ — the redundancy is our cross-check.

3

Geometry

DN / flow area, face-to-face length, and any installation assumptions (upstream straight length, orientation).

4

One worked example

Flow versus pressure loss per stroke at fully stated conditions: medium, absolute pressure, temperature, humidity, and the Nm³ reference basis — or simply mass flow in kg/h, which needs no basis at all.

5

The formula, named

Which sizing relation produced the example — IEC 60534-2-1, a ζ-based relation, or a house formula written out. Named, not implied.

6

Tolerance

The uncertainty band on ζ/Kv (± %). A coefficient without a tolerance cannot carry a guarantee.

7

Where it comes from

Who produced this dataset and from what: a supplier submission, our own digitisation of your published sheet, or a third-party source — plus the document it traces back to. A number without a origin cannot be defended three years later.

8

Characteristic basis

Confirm the curve is the inherent one, measured at constant Δp. An installed curve belongs to somebody else's pipework and cannot be recomputed for our site.

Tier 2 · welcome, not blocking

leakage class at 0 % stroke (EN 60534-4) minimum controllable stroke positioner hysteresis / dead band xT / choked-flow data for high-Δp duty

Submit: fill in, self-check, send

Step 1 — pick the medium class. The coefficients are the same either way (ζ is dimensionless); what changes is which conventions are mandatory.

Step 2 — fill the template below (it is pre-filled with the fictional DN200 valve so the checker demonstrably passes; replace every value with your own). Flows in kg/h are always unambiguous; Nm³/h is accepted only together with the declared basis. Step 3 — run the self-check: it is the same set of checks we run on receipt, so anything red here becomes a clarification round later. Step 4 — return the JSON with your quotation, through the inquiry that pointed you to this page.

How we verify it

Before a valve enters evaluation we rebuild your worked example from your own coefficients and conventions. The example is the checksum: if we cannot reproduce it, we cannot reproduce anything else either.

your dataset ζ · Kv · geometry · conventions we recompute your worked example deviation vs your table ≤ 2 % · accepted valve enters evaluation > 2 % · clarification one question round, then retry

Three consistency checks sit inside that recomputation:

  1. ζ ↔ Kv agree — for the same reference area, ζ ≈ 1.6·10⁹ · D⁴ / Kv² (D in m, Kv in m³/h) must hold across the stroke range.
  2. The example follows √Δp — subcritical gas flow at fixed stroke scales with the square root of the pressure loss; rows that do not are typing errors or an unstated correction.
  3. The physical constant closes — flow, coefficient and stated conditions must reproduce each other through the named formula, not just internally.

Conventions that silently change the numbers

These are the ambiguities we most often have to write back about. Stating them costs one line each; leaving them out changes results by more than typical differences between competing valves.

conventionplease stateeffect when unstated
Nm³ reference basis0 °C / 1.01325 bar, or 20 °C, or other ≈ 7 % on density and flow
ζ reference areanominal bore or seat area up to several % near full stroke
pressure basis of the exampleabsolute upstream pressure Qₙ ∝ √p — 1.49 vs 1.7 bar(a) ≈ 6.8 %
gas-sizing formula variantthe exact relation and constants 1 – 3 % between common variants
humidity basisdry air or a stated RH up to ≈ 1 % on density

Sources

  1. IEC 60534-2-1 — Industrial-process control valves: flow capacity, sizing equations for fluid flow.
  2. EN 60534-4 — Inspection and routine testing (seat leakage classes).
  3. VDI/VDE 2173 — Fluidic characteristic quantities of control valves.
Owner René de Ren — R&D lab
Demo data fictional DN200 valve, generated for illustration on 2026-08-04
Contact R&D lab · wbd-rd.nl