Eleven flow measurement technologies, compared across the twelve criteria that actually decide a specification. Use this when you need to narrow the field quickly, then confirm the detail on the individual product page.
| Technology | Best for | Conductivity | Viscosity limit | Typical accuracy | Turndown | Straight run (up / down) | Pressure loss | Moving parts | Main wear item | Typical application | Main limitation |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Electromagnetic | Conductive liquids, slurries | Yes, 5 µS/cm min | No practical limit | ±0.5% of reading | 20:1 | 5D / 3D | None (full bore) | None | Liner + electrodes | Billing, wastewater, slurry | Will not measure non-conductive fluids |
| Clamp-on ultrasonic | Clean single-phase liquids | Not required | Moderate | ±1% of reading | 40:1 | 10D / 5D | None | None | Couplant, mounts | Retrofit, large mains, surveys | Needs full pipe and measurable wall |
| Vortex | Steam, gas, low-viscosity liquids | Not required | Low only | ±1.0% of reading | 20:1 | 15–30D / 5D | Low | None | None significant | Steam, compressed air | Signal dies below min Reynolds number |
| Liquid turbine | Clean low-viscosity liquids | Not required | <10 cSt | ±0.5% of reading | 10:1 | 10D / 5D | Very low | Rotor + bearing | Bearing | Water, diesel, batching | Bearing wear shifts K-factor |
| Gas turbine | Natural gas, LPG, air | Not required | Low only | ±1.0% of reading | 10:1 | 10D / 5D | Low | Rotor + bearing | Bearing | Gas custody transfer | Needs T/P compensation |
| Thermal mass | Clean dry gas | Not required | Low | ±1–2% of reading | 100:1 | 15–20D / 5D | Very low | Heated sensor | Sensor coating | Compressed air, biogas | Coating drifts reading silently |
| Roots / rotary | Natural gas, town gas | Not required | Low | ±0.5–1% of reading | 100:1 | Short | Low, constant | Rotors | Rotors, seals | Low-pressure gas distribution | Not for dirty or wet gas |
| Orifice / DP | Steam, gas, liquids | Not required | Any | ±1–2% of reading | 3–4:1 | 20–40D / 5D | High, permanent | None | Plate erosion | Steam, high temperature | Permanent pressure loss; blocked impulse lines |
| Balanced orifice | Steam, gas, liquids | Not required | Any | ±0.5% of reading | 5:1 | 5–10D / 5D | Moderate | None | Plate erosion | Retrofit steam metering | Still a DP device with taps |
| Precession vortex | Natural gas, LPG, coal gas | Not required | Low | ±1.5% of reading | 30:1 | 10D / 5D | Low | None | None significant | Gas with built-in compensation | Not for steam or dirty gas |
| Coriolis | Liquids, gases, slurries | Not required | Any | ±0.1–0.2% of reading | 100:1 | None | Low to moderate | Vibrating tubes | Tube fatigue | Custody transfer, dosing | Cost; limited to smaller sizes |
How to read this table
Start with the conductivity column. If the fluid does not conduct, three of these technologies disappear immediately and the shortlist writes itself.
Then look at viscosity. Above about 10 cSt the turbine options fall away, and above 20 cSt most of the velocity-based technologies do too.
Then look at straight run. On a retrofit or a skid with fixed dimensions, the piping constraint usually eliminates more options than the accuracy requirement does.
Accuracy comes last, because it is the specification people argue about most and the one that matters least once the installation is wrong. A plus or minus 0.5 percent meter installed after a close-coupled elbow will not hold half a percent.
Two things this table cannot tell you
Cost. The spread between the cheapest and most expensive option for the same line size is roughly ten to one, and it depends on materials, pressure rating and connection standard as much as on technology. Ask for a quotation on the specific duty.
Whether your installation will let the meter perform. Every accuracy figure above assumes the straight run, the full pipe and the grounding are all correct. Those are installation conditions, not product features.
Checking a specific duty
If you have the flow, the bore and the fluid, the flow rate and pipe velocity calculator will tell you whether the resulting velocity sits in a sensible band before you go any further.
For the decision process in order, see how to choose a flow meter. For the accuracy figures themselves and what they mean at low flow, see flow meter accuracy explained.