Every flow meter datasheet leads with a percentage. Almost every buyer misreads it. The difference between “plus or minus 0.5 percent of reading” and “plus or minus 0.5 percent of full scale” is the difference between a meter that holds accuracy across its range and one that quietly falls apart at low flow.
Of reading, or of span
Percent of reading means the error scales with the actual flow. At 10 m3/h a 0.5 percent of reading meter is within 0.05 m3/h, and at 100 m3/h it is within 0.5 m3/h.
Percent of full scale means the error is fixed to the meter’s maximum. A 0.5 percent of span meter on a 100 m3/h range is within 0.5 m3/h everywhere, including at 10 m3/h, where that is a 5 percent error on the actual value.
Same number on the datasheet. Ten times the real error at low flow. If you are comparing two meters and one quotes reading while the other quotes span, you are not comparing the same thing.
Turndown is where accuracy lives or dies
Every technology has a range over which the stated accuracy holds, and a floor below which it does not. That ratio is the turndown.
A 10:1 turndown on a meter spanning 1 to 10 m3/h means accuracy from 1 m3/h upward. Below 1 m3/h the meter still displays a number. It is just no longer the number you paid for.
This is the single most common sizing mistake we see: a meter chosen for its maximum flow, installed on a line that normally runs at a quarter of it. The meter is not faulty. It is being used outside the part of its curve where the datasheet applies.
Repeatability is not accuracy
Repeatability is how closely the meter returns to the same reading for the same flow. Accuracy is how close that reading is to the true value.
A meter can be highly repeatable and still wrong, if it was calibrated badly or its K-factor was entered incorrectly. That is actually the easier fault to fix. The opposite case, accurate on average but noisy reading to reading, is harder, because you cannot average your way out of it on a batch process.
For control loops and batching, repeatability usually matters more than absolute accuracy. For billing and custody transfer, absolute accuracy is the whole point.
What accuracy do you actually need
| Application | Sensible target | Why |
|---|---|---|
| Plant monitoring, trend only | ±2% of span | Nobody acts on the absolute value |
| Compressed air accounting | ±2–3% of reading | The number drives energy decisions, not invoices |
| Batching and dosing | ±0.5% of reading | Repeatability converts directly into product cost |
| Utility billing | ±1% of reading | Money changes hands |
| Custody transfer | ±0.25–0.5% of reading | Contractual, and usually audited |
We regularly see projects specify 0.5 percent for a cooling water line where nobody will ever look at the absolute value. That is money spent on a specification nobody needed.
Calibration: what the certificate actually proves
A factory calibration certificate proves the meter matched a reference at the time of test, on the fluid and installation used for that test. It does not guarantee the same performance after six months in your line.
Three things move a meter away from its certificate: installation effects (straight run, profile), fluid effects (viscosity, coating, entrained air), and wear (bearings, electrodes). The first is present from day one, the other two accumulate.
This is why in-situ verification matters on critical duties. A clamp-on ultrasonic meter strapped next to an installed meter is the cheapest way to find out whether the installed one is still telling the truth, and it takes an hour rather than a shutdown.
The practical rule
Ask three questions before accepting any accuracy figure. Is it of reading or of span? Over what turndown does it hold? And under what installation conditions was it established?
If a supplier cannot answer all three, the percentage on the datasheet is marketing rather than engineering.
Related reading
If you are still choosing between technologies, our flow meter selection guide works through the decision by fluid type. To check whether your line velocity is sensible before you size anything, use the flow rate and pipe velocity calculator.
Comparing technologies side by side? The flow meter technology comparison sets 11 types against 12 selection criteria in one table.