How to Choose an Industrial Flowmeter for Liquids and Gases – Complete Guide

9 May 2026

Choosing an industrial flowmeter is one of the instrumentation decisions with the greatest impact on a process. A wrong choice does not simply mean an inaccurate measurement. It means incorrect energy bills, undetected raw material losses, unstable process control, or – in fiscal metering applications – commercial disputes.

The real challenge is that at least seven different flow measurement technologies are available on the market, each with clear advantages and equally clear limitations. There is no universal technology. An excellent electromagnetic flowmeter for water becomes completely unusable on gases or hydrocarbons. A thermal mass flowmeter that works perfectly on compressed air will produce serious errors on a liquid.

This guide explains each technology, compares their applicability for liquids versus gases, and provides a clear selection process.

What a Flowmeter Measures

A flowmeter quantifies the amount of fluid passing through a pipe cross-section per unit of time. There are two fundamental measurement types:

Volumetric flow – expressed in m³/h, l/h, l/min. Measures the volume of fluid without accounting for density. If temperature or pressure varies, the volume changes for the same mass of fluid.

Mass flow – expressed in kg/h, t/h. Directly measures the mass of fluid. Not affected by variations in temperature, pressure, or density.

The choice between volumetric and mass flow depends on the application. Chemical processes, dosing, and fiscal gas transfer require mass flow. Monitoring applications, water supply, or irrigation work with volumetric flow.

The 7 Main Flow Measurement Technologies

1. Electromagnetic Flowmeter (Magmeter)

Operating principle: A conductive fluid flows through a magnetic field generated by external electromagnets. By electromagnetic induction (Faraday’s law), a voltage is generated proportional to the flow velocity.

Where it works well:

Drinking water, wastewater, sludge
Acid and alkali solutions
Electrically conductive pastes and slurries
Conductive food liquids

Clear limitations:

Does not work on non-conductive fluids – hydrocarbons (oil, diesel, petrol), pure oils, ultrapure water
Cannot measure gases or steam
Requires fully filled pipe – does not function in partially filled pipes
Requires minimum fluid conductivity (≥5 µS/cm in standard versions)

Typical accuracy: ±0.2–0.5% of measured value
Typical turndown: 30:1 to 1000:1
Installation requirements: 5–10D upstream, 2–5D downstream

2. Coriolis Flowmeter

Operating principle: Fluid passes through one or two vibrating tubes oscillating at their natural frequency. The Coriolis force generated by the moving mass causes a tube twist proportional to mass flow. The same principle allows direct density measurement.

Where it works well:

Applications where accuracy is critical: fiscal transfer, chemical dosing, product recipes
Liquids of any type – water, hydrocarbons, acids, pastes, syrups
Gases – including process gases, CO₂, liquefied gases
Fluids with variable density or viscosity
Applications where both mass flow and density are needed simultaneously

Clear limitations:

The most expensive flowmeter type
Higher pressure drop compared to non-invasive technologies
Large gas bubbles in liquid can disrupt tube vibration
Not economically recommended for very large pipe diameters

Typical accuracy: ±0.1–0.5% of measured value (among the best on the market)
Typical turndown: 100:1 and above
Installation requirements: no strict straight pipe run requirements; avoid installation near strong vibration sources

3. Vortex Flowmeter

Operating principle: A bluff body placed in the flow path generates alternating vortices (von Kármán effect). The vortex frequency is directly proportional to the flow velocity.

Where it works well:

Steam (saturated and superheated) – one of the most common applications
Clean utility gases (compressed air, nitrogen, CO₂)
Clean, low-viscosity liquids

Clear limitations:

Does not work at low flow rates – below a critical minimum (Reynolds number threshold), vortices stop forming and measurement becomes impossible
Not recommended for fluids with particles or viscous fluids (build-up on bluff body)
Sensitive to external vibrations, which can be interpreted as false vortices
Requires significant straight pipe run: 15–20D upstream after bends in different planes, 5D downstream

Typical accuracy: ±0.5–1% of measured value
Typical turndown: 20:1 to 40:1
Installation requirements: 15–20D upstream, 5D downstream

4. Ultrasonic Flowmeter

Two variants with different operating principles:

Transit time: Two piezoelectric transducers alternately emit ultrasonic pulses in and against the direction of flow. The difference in propagation time is proportional to the fluid velocity.

Works on clean liquids or liquids with fine particles
Does not work well if gas bubbles are present – ultrasound is scattered
Available in clamp-on version (external mounting on pipe, no pipe cutting required)
Suitable for large pipes and retrofit without process shutdown

Doppler: Emits ultrasonic waves that reflect off particles or bubbles in the fluid.

Works on liquids with suspended particles or bubbles
Less accurate than transit time
Not recommended for clean liquids

Where it works well:

Water, wastewater, clean chemical liquids (transit time)
Liquids with suspensions or bubbles (Doppler)
Large pipes where in-line installation would be costly
Retrofit or temporary monitoring situations (clamp-on)
Clean gases – multi-path models perform well on gases

Clear limitations:

Gas bubbles in liquid destroy transit time measurement
Clamp-on accuracy is lower than in-line mounting
Pipe inner wall must provide good acoustic contact with transducers

Typical accuracy: ±0.5–2% (in-line), ±1–3% (clamp-on)
Typical turndown: 30:1 to 100:1
Installation requirements: 10D upstream, 5D downstream (in-line); 15–30D upstream (clamp-on)

5. Thermal Mass Flowmeter

Operating principle: Two temperature sensors and a heating element measure the heat transfer of the passing gas. The heat flux is proportional to mass flow.

Where it works well:

Clean or lightly contaminated gases: compressed air, nitrogen, CO₂, biogas, natural gas, combustion air
Low gas flow rates where other technologies lose accuracy
Applications requiring mass flow without external temperature and pressure compensation
Wastewater treatment plants (aeration air measurement)

Clear limitations:

Does not work on liquids in the standard configuration
Sensitive to gas composition – if the gas changes (e.g., variable mixture), the original calibration becomes inaccurate
Moisture or condensate on the sensor produces significant errors and can damage the sensor
Not recommended for corrosive or saturated (wet) gases
Higher pressure drop in inline versions at high flow rates

Typical accuracy: ±1–2% of measured value
Typical turndown: 10:1 to 100:1
Installation requirements: 10–15D upstream, 5D downstream

6. Positive Displacement (PD) Flowmeter

Operating principle: Precision mechanisms (lobe rotors, pistons, oval gears) capture exact volumes of fluid and transfer them from inlet to outlet. The number of rotations is directly proportional to the total volume of fluid passed.

Where it works well:

Viscous fluids: oils, lubricants, heavy fuels, syrups, paints
Fiscal transfer and fluid custody applications (fuels, lubricants)
Low flow rates with high accuracy requirements
Remote locations without electrical power (mechanical versions)

Clear limitations:

Requires clean fluid – particles larger than 100 µm damage precision components; filtration is mandatory
Not recommended for abrasive fluids
Pressure drop increases with viscous fluids and high flow rates
Moving parts are subject to wear and require periodic maintenance
Not used on gases in standard liquid versions

Typical accuracy: ±0.1–0.5% of measured value
Typical turndown: 15:1 to 100:1 (higher with high-viscosity fluids)
Installation requirements: no straight pipe run requirements

7. Differential Pressure Flowmeter (DP – Orifice Plate, Venturi, Pitot)

Operating principle: A calibrated flow restriction (orifice, Venturi, Pitot tube) creates a pressure differential proportional to the square of the flow velocity. A DP transmitter measures this differential and calculates flow.

Where it works well:

Standard liquid, gas, and steam applications
Large pipelines where other technologies would be too costly
Situations where simplicity and low cost take priority over maximum accuracy

Clear limitations:

Permanent pressure loss (higher for orifice plate, lower for Venturi)
Lower accuracy compared to modern technologies (±0.5–2%)
Limited turndown: typically 3:1 to 5:1 on a single range
Sensitive to fluid density variations (requires PT compensation)
Orifice plate can be eroded by fluids with abrasive particles

Typical accuracy: ±0.5–2% of measured value
Typical turndown: 3:1 to 5:1
Installation requirements: 20–40D upstream, 5D downstream (varies by type)

Direct Comparison: Liquids vs. Gases

Liquids – Which Technology Works

Technology	Clean Liquids	Viscous Liquids	Liquids with Particles	Hydrocarbons	Steam
Electromagnetic	✓	✓	✓	✗	✗
Coriolis	✓	✓	✓	✓	✗
Vortex	✓	✗	✗	✓	✓
Ultrasonic transit time	✓	✓	partial	✓	✗
Ultrasonic Doppler	partial	partial	✓	partial	✗
Positive Displacement	✓	✓ (excellent)	✗	✓	✗
DP (orifice/Venturi)	✓	✓	partial	✓	✓

Gases – Which Technology Works

Technology	Clean Gases	Wet Gases	Corrosive Gases	Steam	Direct Mass Flow
Thermal mass	✓ (excellent)	✗	partial	✗	✓
Coriolis	✓	✓	✓	✗	✓
Vortex	✓	partial	partial	✓	with compensation
Ultrasonic	✓	partial	partial	✗	with compensation
DP (orifice/Venturi)	✓	✓	✓	✓	with compensation
Electromagnetic	✗	✗	✗	✗	✗
Positive Displacement	✗	✗	✗	✗	✗

How to Select Correctly – Practical Steps

Step 1 – Define the fluid

Is it a liquid or a gas? (immediately eliminates several options)
Is it electrically conductive? (no → electromagnetic excluded)
Does it contain particles or bubbles? (influences ultrasonic selection)
Is it viscous? (>10 cP → positive displacement or Coriolis)
Is it corrosive? (special materials and coatings)
Is the composition constant or variable? (critical for thermal and Coriolis on gases)

Step 2 – Define process conditions

Maximum and minimum operating pressure
Maximum and minimum operating temperature
Flow range: minimum, normal, maximum (required turndown)
Flow variations: frequent or is the process stable?

Step 3 – Define accuracy requirement

General monitoring: ±1–2% is sufficient → DP, clamp-on ultrasonic
Process control: ±0.5% → electromagnetic, vortex, in-line ultrasonic
Fiscal metering or critical dosing: ±0.1–0.2% → Coriolis, positive displacement

Step 4 – Evaluate installation conditions

Is there enough space for the required straight pipe run?
Can the process be interrupted for installation?
Are there vibrations or intense electromagnetic fields present?
What is the pipe diameter?

Step 5 – Define output and integration requirements

Required signal: 4-20 mA, HART, Modbus, Profibus, Profinet?
Are additional functions needed: temperature, density, mass flow calculation?
Are certifications required: ATEX, IECEx, SIL?

Step 6 – Evaluate total cost of ownership

Initial cost is only part of the picture. Total cost includes:

Purchase price
Installation cost (including pipe modifications)
Calibration frequency and cost
Pressure drop (energy consumed over service life)
Spare parts and maintenance costs

Common Mistakes in Flowmeter Selection

Mistake 1 – Selecting based on purchase price only. A DP flowmeter with an orifice plate costs significantly less than an electromagnetic or Coriolis meter. But the permanent pressure drop consumes energy every day. On a 50 kW pump running 8,000 h/year, even a few millibar of pressure loss translates into significant annual costs.

Mistake 2 – Ignoring fluid properties. The most frequent error: installing an electromagnetic flowmeter on a non-conductive fluid (oil, diesel, petrol). The instrument will not function at all or will produce completely random readings.

Mistake 3 – Underestimating the required turndown. If the process requires correct measurement at both 10% and 100% of nominal flow, a turndown of at least 10:1 is needed. A standard DP flowmeter has a turndown of 3:1–5:1, making it unsuitable for processes with large flow variations.

Mistake 4 – Ignoring installation conditions. A vortex flowmeter installed 2D upstream of a bend will produce errors of 5–15%, regardless of sensor quality. The accuracy declared by the manufacturer is valid only when installation requirements are met.

Mistake 5 – Choosing a thermal flowmeter for gases with variable composition. If gas composition changes (e.g., biogas with variable CH₄/CO₂ content), a thermal flowmeter calibrated for a fixed composition will introduce errors proportional to the deviation from calibration conditions. The solution is a Coriolis meter or an ultrasonic with composition compensation.

Mistake 6 – Using a vortex meter at low flow rates. Below a minimum flow specific to each model, von Kármán vortices no longer form and the instrument cannot make any measurement. If the process involves periods of zero or near-zero flow, a vortex meter is not the right choice.

Practical Examples from Industry

Water treatment plant – DN300. Fluid: drinking water, good conductivity. Decision: electromagnetic flowmeter. No moving parts, negligible pressure drop, ±0.2% accuracy, 1000:1 turndown.

98% sulphuric acid dosing line – DN25. Fluid: concentrated sulphuric acid, corrosive, small and critical flow. Decision: Coriolis flowmeter in Hastelloy. ±0.1% accuracy, direct mass measurement, adequate chemical resistance. The high cost is justified by the critical nature of the dosing application.

Compressed air metering for energy balance – DN50. Fluid: compressed air, clean gas. Decision: thermal mass flowmeter. Direct mass flow without PT compensation, affordable cost, wide turndown.

Diesel fuel custody metering – tank to pump – DN40. Fluid: diesel fuel, hydrocarbon, medium viscosity. Decision: positive displacement flowmeter (oval gear). Fiscal metering accuracy ±0.1%, no straight pipe run requirements.

Saturated steam measurement – DN100. Fluid: saturated steam at 180°C. Decision: vortex flowmeter with temperature and pressure compensation. High temperature excludes standard clamp-on ultrasonics. Vortex is the reference technology for steam in industrial applications.

Conclusion

Choosing an industrial flowmeter is not a simple decision. It depends on the fluid, process conditions, accuracy requirements, available installation space, and total cost of ownership.

The basic rule:

Water and conductive liquids → electromagnetic
Clean gases → thermal mass or vortex
Steam → vortex
Maximum accuracy or mass flow → Coriolis
Viscous liquids or fiscal metering → positive displacement
Retrofit without shutdown or large pipes → ultrasonic clamp-on
Minimum cost on large pipelines → DP (orifice/Venturi)

Every application has its optimal solution. Investing in the right selection from the beginning eliminates measurement problems, rework, and hidden costs over the service life of the installation.

For technical advice on sizing and selecting the right flowmeter for your process, the DIVINOV ENGINEERING team provides specialised support – from application analysis to equipment delivery and commissioning.