Flow Meter Accuracy Calculation: Add Terms or Choose a Range?
Flow Measurement Guide
A flow meter accuracy calculation starts with the wording beside the percentage. A reading term, a full-scale term and a specification that changes at a stated flow are different instructions. Identify the instruction first, then calculate the stated tolerance at the flow you need.
1. Copy the complete specification
Record the percentage, its basis, the applicable flow interval and any conditions. % of reading changes with the reading; % of full scale uses the full-scale value defined for that specification. Bronkhorst illustrates why the latter becomes a larger percentage of the measured flow at low flow. Bronkhorst: accuracy and repeatability
Do not silently substitute the display's configured upper limit for the manufacturer's definition of full scale. Keep the specified units and, for gas volume, the reference conditions consistent. If a specification says setpoint, use setpoint; it is not automatically interchangeable with a measured reading.
2. Decide whether the terms are added or selected
Two percentages on a datasheet do not always mean addition.
| Wording pattern | Calculation instruction |
|---|---|
| ±a% of reading | Multiply the reading by a/100. |
| ±b% of full scale | Multiply the specified full-scale value by b/100. |
| ±(a% of reading + b% of full scale) | Calculate both terms and add their magnitudes. |
| One specification above a threshold; another below it | Select the applicable interval, then use its formula. Do not add the two rows. |
These patterns occur in real specifications. KOBOLD lists a reading-plus-full-scale expression for its MIM meter. Brooks gives an MFC example with a setpoint-based specification from 20–100% of full scale and a full-scale specification below 20%. The Brooks example concerns a controller and setpoint; it demonstrates a conditional specification, not a reading-based claim for every flow meter. KOBOLD · Brooks Instrument
3. Calculate at the same operating flow
The following specifications are hypothetical teaching examples, not ratings of those manufacturers' products. Assume full scale is 100 L/min and every example is valid from 10 to 100 L/min under the same stated conditions.
- A: ±1% of reading.
- B: ±0.5% of full scale.
- C: ±(0.5% of reading + 0.1% of full scale).
- D: ±1% of reading at 20–100 L/min; ±0.2% of full scale from 10 to below 20 L/min.
Each cell gives ±L/min (equivalent ±% of reading).
| Reading | A | B | C | D |
|---|---|---|---|---|
| 10 L/min | 0.10 (1%) | 0.50 (5%) | 0.15 (1.5%) | 0.20 (2%) |
| 20 L/min | 0.20 (1%) | 0.50 (2.5%) | 0.20 (1%) | 0.20 (1%) |
| 50 L/min | 0.50 (1%) | 0.50 (1%) | 0.35 (0.7%) | 0.50 (1%) |
| 100 L/min | 1.00 (1%) | 0.50 (0.5%) | 0.60 (0.6%) | 1.00 (1%) |
At 10 L/min, C gives 0.005 × 10 + 0.001 × 100 = 0.15 L/min. For D, 10 is below the threshold, so use only 0.002 × 100 = 0.20 L/min. Adding D's two branches would incorrectly produce 0.30 L/min.
To express a calculated tolerance E as a percentage of a positive reading Q, calculate 100 × E/Q. This conversion is undefined at zero flow. At exactly 20 L/min, D uses the interval that includes 20; always check how the actual datasheet assigns its boundary.
4. Use the result within its limits
Repeat the calculation at your minimum, normal and maximum flow. In the example, A has a smaller stated tolerance than B at 10 L/min, while B has the smaller tolerance at 100 L/min. A headline percentage alone cannot establish the better match.
Do not extend a formula below its specified measuring range. Spirax Sarco explains how turndown limits the interval over which stated performance applies. Principles of flowmetering
These calculations translate specification wording. They do not establish actual measurement error, calibration conformity or the uncertainty of an installed measuring system. Installation, fluid conditions, configuration and the relevant reference measurement still require their own assessment.
For a separate comparison of measuring principles and application requirements, see Coriolis vs electromagnetic flowmeters.