Flowmeter Specifications: How to Read a Datasheet the Right Way

Two flowmeter datasheets can both promise "1% accuracy" and describe genuinely different instruments; one can be four times better than the other at the flow rate you actually run. The gap is not a manufacturing defect or a marketing lie. It is a single, quietly important phrase buried in the small print: accuracy of reading, or accuracy of full scale. Get that distinction wrong and every other figure on the sheet, turndown, repeatability, pressure rating, becomes harder to trust.

Nixon Flowmeters has read flowmeter datasheets for a living since 1982, across turbine, magnetic, ultrasonic and variable area technologies, and for Tecfluid as well as our own NT and NTG ranges. This article works through a real flowmeter datasheet line by line: the accuracy convention that catches most buyers out, the difference between accuracy and repeatability, what turndown actually promises, why reference conditions matter, and the questions a datasheet cannot answer at all. By the end you should be able to read one like a specialist rather than take the headline number on trust.

The Accuracy Trap: Percent of Reading vs Percent of Full Scale

Every flowmeter accuracy figure is expressed against one of two references, and datasheets rarely spell out which one loudly enough. Accuracy of full scale (also written "of span" or "FS") is a fixed error, a percentage of the meter's maximum rated flow, that stays the same absolute size no matter what the actual flow is at the time. Accuracy of reading (also "of rate") is a percentage of whatever the flow happens to be right now, so the error shrinks and grows in proportion to the flow itself.

A meter quoted at 1% of full scale and a meter quoted at 1% of reading are only equally accurate at one single point: maximum flow. Everywhere else in the range, the full-scale meter gets progressively worse in relative terms as flow drops, while the reading meter holds its 1% steady throughout its linear range.

The Arithmetic in One Table

Take two meters, both rated 0 to 100 m³/h and both marked "1% accuracy," one quoted of full scale and one of reading:

At a quarter of the range, the full-scale meter is quietly a 4%-of-reading instrument, four times worse than its headline figure suggests, while the reading-based meter is exactly as accurate as it was at maximum flow. Neither convention is dishonest by itself; both are standard industry practice, and both appear on genuine, well-engineered datasheets. The problem is only ever a buyer who assumes "1% accuracy" means the same thing on every sheet.

Our turbine flowmeters are typically quoted of reading, which is why the figure holds steady across the meter's linear range. Tecfluid's variable area flowmeters, which Nixon supply as UK agent, are a good honest example of the other convention: accuracy quoted of full scale, simple and transparent, but a number that needs interpreting correctly at the bottom of the scale rather than read at face value.

Accuracy vs Repeatability: Two Different Promises

Accuracy and repeatability answer different questions, and a datasheet that only shouts about one is not giving you the full picture. Accuracy is closeness to the true value, traceable back to a calibration reference. Repeatability is closeness to the meter's own previous readings under the same conditions; it says nothing about whether that consistent value is actually correct.

For custody transfer, fiscal metering or anywhere a reading has commercial or legal consequences, absolute accuracy is what matters. For batching, blending and most process control, repeatability is usually the number worth chasing, because you are matching today's batch to yesterday's, not to an external reference standard. A well-built turbine meter typically holds repeatability better than 0.1%, considerably tighter than its quoted accuracy figure, which is exactly why it earns its keep in batching duties even where absolute accuracy is more modest.

Turndown and Rangeability: Where the Accuracy Actually Holds

Turndown (sometimes "rangeability") is the ratio between a meter's maximum and minimum flow within its calibrated, linear range, commonly quoted as 10:1 or 20:1. It sounds like a straightforward promise of flexibility. The question a buyer should always ask next is: over what part of that range does the stated accuracy actually hold?

A wide turndown demonstrated only at the top of the range, with accuracy quoted at maximum flow and left silent on performance near minimum flow, is a brochure trick that costs nothing to print and a great deal to discover in service. Ask for accuracy across the full stated turndown, not just at the flow rate that makes the datasheet look best, and size the application so normal flow sits comfortably within the range rather than pinned against either edge.

Reference Conditions: What the Number Was Actually Calibrated Against

Every accuracy figure was established under specific conditions: a calibration fluid, usually water, at a stated viscosity and temperature. A water calibration does not automatically transfer to your fuel oil, your chemical blend or your compressed gas, and a datasheet that does not state the calibration fluid is one worth questioning before you trust the headline number for a different process fluid.

Viscosity sensitivity varies sharply by technology. Turbine meters are viscosity-sensitive: rotor drag changes with viscosity, and a fuel that thickens in cold weather can drift the meter low against its original calibration. Variable area meters are calibrated against a specific fluid density and viscosity too, which is why a genuine VA datasheet states the calibration fluid explicitly. Magnetic flowmeters, by contrast, measure via Faraday's law of induction and are unaffected by viscosity or density, which is one reason they are the honest, low-drama choice for variable or poorly characterised liquids.

Pressure, Temperature and Materials: The Lines Buyers Skim

Two pressure figures appear on most datasheets and they are not the same thing. Pressure rating is the meter body's rated tolerance to line pressure, a structural limit. Pressure drop (or headloss) is the pressure the meter itself removes from the flow as fluid passes through it, which rises with flow rate and matters for pumping cost and downstream pressure. Confusing the two is an easy, avoidable mistake.

Wetted materials determine chemical compatibility with your process fluid: stainless steel, bronze, PTFE seals and similar all behave differently against acids, solvents and abrasives, and the wrong choice fails quietly over months rather than obviously on day one. Temperature limits also split in two: ambient temperature limits govern the electronics and housing, while process temperature limits govern the wetted parts and seals in contact with the fluid itself. A datasheet listing only one is not telling you the whole story.

Outputs and Approvals

Flowmeter outputs range from a raw pulse train, which needs the meter's K-factor to convert into a flow rate, through 4-20 mA analogue, HART overlay on the same 4-20 mA loop, to fully digital Modbus. The right choice depends entirely on what is reading the signal: a simple local indicator, a PLC, or a plant-wide SCADA system.

Approvals matter where the installation or the reading demands them. ATEX markings are required for hazardous-area installations handling flammable liquids or gases. Metrological approvals such as MID become important specifically when a reading has commercial consequences, custody transfer and billing being the obvious cases, and are worth confirming explicitly rather than assuming they are implied by a general accuracy claim.

What a Datasheet Cannot Tell You

A datasheet is tested under ideal conditions, and few real installations are ideal. It cannot tell you how sensitive the meter will be to a short straight run or a distorted velocity profile in your particular pipework; it cannot fully capture how your actual process fluid, with its entrained air, temperature swings or trace contamination, will behave compared with the clean calibration fluid used to establish the figures; and it rarely gives a guaranteed service life for wear parts such as bearings and seals, because that depends on your duty, not the meter's design alone.

These are genuinely important gaps, not small print to be waved away. If a meter is already installed and behaving oddly, our guide to troubleshooting flow meter errors covers the common symptoms and causes. If you are still choosing a technology, our comparison of turbine, magnetic and ultrasonic flowmeters and our detailed look at turbine accuracy and turndown are the next places to look.

Frequently Asked Questions

What is the difference between accuracy of reading and accuracy of full scale?

Accuracy of reading is a fixed percentage of the actual flow at that moment, so it holds steady across the meter's linear range. Accuracy of full scale is a fixed absolute error based on the meter's maximum rated flow, so it becomes progressively worse in relative terms as flow drops below maximum.

What is turndown on a flowmeter?

Turndown, or rangeability, is the ratio between a meter's maximum and minimum flow within its calibrated, linear range, commonly 10:1 or 20:1. The critical question is whether the stated accuracy holds across that whole range or only at the top of it.

What does repeatability mean on a flowmeter datasheet?

Repeatability is how closely a meter reproduces the same reading under the same flow conditions, run after run. It is a measure of consistency with itself, not closeness to a true reference value, which is a separate figure: accuracy.

What outputs do flowmeters have?

Common outputs are a raw pulse train (requiring the K-factor to convert to flow rate), 4-20 mA analogue, HART overlay on the analogue loop, and digital Modbus. The right choice depends on what device or system is reading the signal.

Why do two flowmeters with the same accuracy figure perform differently?

Because the figure can be expressed of reading or of full scale, calibrated against different reference fluids and viscosities, and quoted across different turndown ranges. Two identical-looking headline percentages can describe meaningfully different real-world performance.

Reading the Specification Properly

A datasheet will get you to a shortlist. It will not tell you how a meter behaves on your fluid, in your pipework, at your normal operating flow, and no amount of careful reading changes that; only an application conversation does. Understanding whether a figure is quoted of reading or of full scale, whether turndown accuracy has been demonstrated across the whole range or just at the top, and what the meter was actually calibrated against gets you to a properly informed shortlist rather than a hopeful guess.

If you would like a second pair of eyes on a specification, the Nixon team reads these documents for a living and is happy to help.