Definitions
Both instruments deliver true mass flow rate on a single 4-20 mA loop, but through different physics. The mechanism determines where each meter wins. On the quote desk we see this confusion show up weekly: a buyer asks for “Coriolis quality” on a low-pressure biogas duty where thermal mass is the right tool, or specs thermal mass on a 7 MPa hydrogen line where Coriolis is the only safe answer.
Thermal Mass Flow Meter
Measures mass flow by the heat the gas carries away from a heated RTD probe. A reference RTD tracks ambient gas temperature; a second RTD is held at a constant temperature differential (CTD) above it. The electrical power required to maintain that differential is proportional to mass flow rate; via gas-specific heat capacity.
Direct output: Nm³/h or kg/h. No external pressure or temperature correction required.
Coriolis Flow Meter
Measures mass flow by the Coriolis force acting on a vibrating tube carrying the fluid. Drive coils oscillate one or two U-tubes at resonance; mass flow causes the inlet leg to twist against the outlet leg. A pair of pickup coils measures the phase delay between the twists; phase delay is directly proportional to mass flow.
Direct output: kg/h (plus density and temperature as secondaries). Fluid-independent.
Formulas.
The measuring equations explain why the two meters behave so differently in service.
Thermal mass (thermal-dispersion method).
Coriolis (phase-delay method).
Two takeaways: thermal mass reading scales with cp; if you change gases, you change calibration. Coriolis scales with tube geometry only; one calibration, any fluid. That single difference drives 80 % of the selection logic below.
Units and Symbols
| Quantity | Symbol | SI unit | Industry units |
|---|---|---|---|
| Mass flow rate | Qm, ṁ | kg/s | kg/h, t/h, lb/min |
| Reference-volume flow | Qv,ref | Nm³/s | Nm³/h (0 °C), m³/h (20 °C, GB/T), SCFM (70 °F, NIST) |
| Specific heat (constant p) | cp | J/(kg·K) | kJ/(kg·K) |
| Density at reference | ρref | kg/m³ | kg/Nm³ |
| Temperature differential | ΔT | K | °C (same magnitude) |
| Phase delay (Coriolis) | Δt | s | μs |
Converting between readings.
Thermal mass meters usually display Nm³/h; Coriolis meters usually display kg/h. Both are mass-equivalent; converting between them requires only the gas density at the chosen reference condition.
Reference conditions matter: 1,000 Nm³/h at IUPAC STP (0 °C) = 1,073 m³/h at GB/T 3102.8 (20 °C) = 588 SCFM at NIST (70 °F). If your procurement contract is in SCFM and your meter displays Nm³/h, the 8 % reference-condition difference can swing invoicing by thousands of dollars a month. Specify the reference condition at order; every ProFlow meter supports switching via HART without factory intervention.
Side-by-Side Comparison
| Attribute | Thermal Mass | Coriolis |
|---|---|---|
| Accuracy (gas, in service) | ±1.0 to ±1.5 % of reference volume | ±0.5 to ±1.0 % of mass |
| Turndown | 1:100 to 1:1200 on one model | 1:20 to 1:50 (density-limited) |
| Low-pressure capability | Typical ±1.0 % floor holds to ~10 kPa gauge | Accuracy degrades below ~300 kPa; signal-noise-limited below 50 kPa |
| Pipe size range | DN10 to DN2000 (insertion above DN200) | DN6 to DN400 (single-tube sensor weight limit) |
| Straight pipe required | 10×DN upstream, 5×DN downstream | None; install after elbow |
| Primary output | Nm³/h direct (billing-ready) | kg/h direct (needs ρref for Nm³/h) |
| Gas-mix tolerance | Requires calibration curve per gas mix | Truly fluid-independent |
| Two-phase flow | Fails; do not use | Degraded but usable |
| Pressure drop at rated flow | Negligible (wall-mounted probe) | ~0.1–0.8 bar typical; varies widely with tube geometry (straight, U, dual-bent) |
| Weight (DN50, dry) | 8–12 kg | 20–35 kg (vendor-dependent; Promass F ≈ 30 kg) |
| Installed cost (DN50 natural gas) | USD 1,200–2,200 | USD 6,000–12,000 |
| Custody-transfer approval | Industrial / light-commercial trade grade | Full trade-grade custody-transfer class including MID Annex MI-002 |
Green cells indicate the winner on each attribute. Source: ProFlow engineering benchmarks against Emerson Micro Motion CMF + Endress+Hauser t-mass specification sheets, Q1 2026.
Decision Tree
7 questions; answer in order
Selection by industry.
Compressed air, nitrogen, oxygen plants.
Thermal mass wins on nearly all duties. Variable-load compressed air mains, VPSA O₂ output monitoring, N₂ blanketing on chemical tanks; all are low-pressure (<1 MPa), variable-flow, and energy-accounting applications. A single DN100 thermal mass covers 50–5,000 Nm³/h on one meter. On actual orders we see plant engineers default to thermal mass here even when the original spec called for Coriolis. Once procurement runs the numbers on calibration cadence (thermal mass: 12–12 months in service; Coriolis: every 12 months on custody), the lifetime cost gap widens further. Standard compressed-air TMF models ship in 5–7 business days from order confirmation. See TMF for compressed air and TMF for oxygen.
Natural gas (low- and medium-pressure distribution).
Split decision. City-gate and regulator-station flow (0.2–0.6 MPa) → thermal mass or Coriolis both work; choose on cost and turndown. Pipeline transmission (>2 MPa, custody transfer) → ultrasonic dominates DN150+ (AGA-9 / ISO 17089) because Coriolis tube-pressure-drop becomes uneconomic at large diameters; orifice plate is still common on legacy lines. Industrial burner feed (0.01–0.05 MPa) → thermal mass only. See our volumetric vs mass guide for why natural gas is almost always billed in Nm³ (mass-equivalent) rather than m³.
Hydrogen.
Coriolis for custody-transfer dispensing in hydrogen fueling stations (70 MPa dispensing targets ±0.5 % mass accuracy under OIML R139 / SAE J2601 / NIST HB-44 §3.39; not AGA-11, which is an ultrasonic-gas standard). Thermal mass for electrolyser output monitoring, industrial reform feed, refinery sub-measurement, and general process flow. Materials matter: both meters need 316L or higher on H₂-embrittlement-risk service per ASME B31.12. See our hydrogen flow measurement guide.
Biogas, landfill gas, flue gas (CEMS).
Thermal mass wins on all three. Low pressure, variable composition (the CH₄ fraction in biogas swings 50–70 %), and heavy particulate make Coriolis impractical. For CEMS (Continuous Emissions Monitoring Systems) applications under EPA Part 75, thermal mass with a calibration curve verified against the actual stack gas composition meets the ±3 % field accuracy requirement. See TMF for flue gas.
Chemical and pharma dosing.
Coriolis; density and mass both matter for reaction stoichiometry, and most liquid dosing duties are below 1:20 turndown where Coriolis shines. Mass fraction on multi-component streams falls out of the density reading for free. On these duties we ship custody-grade Coriolis configurations in 5–7 business days from order confirmation; sanitary-finish versions for pharma add a week for finish-and-passivation.
Worked example: natural gas burner header.
SERVICE DATA
Application: Industrial boiler burner-header, pipeline natural gas 95 % CH₄ balance ethane/N₂.
Line size: DN80 carbon steel.
Operating pressure: 0.05 MPa gauge (low-pressure post-regulator).
Flow range: 80 Nm³/h (burner idle) to 1,200 Nm³/h (peak firing); 1:15 turndown.
Accuracy target: ±2 % for energy-accounting reconciliation.
Budget: < USD 3,500 installed.
Work the decision tree:
Q2 Pressure < 0.3 MPa? → YES; thermal mass.
Result: thermal mass meter, DN80, calibrated for pipeline NG mix, 4–20 mA + HART, direct Nm³/h at GB/T 3102.8 (20 °C) reference. Budget cleared by ~60 %. Specify a flow-straightener sleeve to reduce upstream pipe to 5×DN if layout is tight. See our TMF-NG natural gas meter for the matching model.
In the field, we ship roughly this same configuration eight to ten times a month for boiler-house and reformer-feed duties. Standard configurations leave our factory in 5–7 business days; if you need calibration certificates traced against a specific NG composition (Wobbe-banded), add a week for the cert lab.
Common questions
What is the difference between Coriolis and thermal mass flow meter?
Both read mass flow directly, but the physics is different:
- Coriolis measures the twist of a resonant vibrating tube caused by the Coriolis force on the moving mass inside. Reads mass + density + temperature on the same instrument. Indifferent to gas composition. ±0.1 to ±1.0 % accuracy depending on tier.
- Thermal mass measures the heat transfer from a heated RTD sensor to the gas flowing past it. Heat lost scales with mass flow rate. Calibrated per gas type (one curve per gas mix). ±1.0 to ±1.5 % accuracy at calibrated composition.
Coriolis fits custody-billing, multi-component streams, and any liquid duty. Thermal mass fits compressed air, single-component gas (N₂, O₂, CH₄), and variable-load duty where turndown matters more than absolute accuracy. The choice is rarely accuracy alone: it is accuracy × cost × turndown × install constraint for the specific duty.
What are the drawbacks of using thermal mass?
Four constraints worth knowing before specifying:
- Gas-composition dependence. Calibration is gas-specific. A meter calibrated on pure CH₄ reads 5 to 10 % low on a 90/10 CH₄/CO₂ biogas blend unless recalibrated. Multi-component or drifting streams need either a fresh calibration or a real-time GC pairing.
- Two-phase failure mode. The heated RTD sensor fails outright in liquid contact; droplets cool the sensor (over-reading flow) and sustained liquid damages the element. Knock-out pot required upstream on any wet-gas service.
- Lower accuracy ceiling. ±1.0 % of rate is the catalog floor. Custody-transfer regimes demanding ±0.5 % or tighter rule it out.
- Straight-pipe requirement. 10×DN upstream + 5×DN downstream typical, reducible with a flow-straightener sleeve. Coriolis needs zero straight pipe.
For dry single-component gas with variable load and accuracy budget ±1 % or looser, thermal mass is the cost-effective answer. For everything else, look at Coriolis or a different principle entirely.
How accurate is a thermal mass flow meter?
Catalog accuracy is ±1.0 to ±1.5 % of rate at calibrated gas composition and reference flow conditions. Field accuracy compounds three contributors:
- Calibration-gas mismatch. A 10 % CO₂ excursion on a CH₄-calibrated meter shifts the reading 5 to 8 %. Stable composition holds catalog accuracy; drifting composition does not.
- Installed straight-run. Without the 10×DN upstream / 5×DN downstream pipe, velocity-profile distortion adds 2 to 4 % systematic error. A flow-straightener sleeve recovers most of that.
- Long-term drift. RTD aging adds ~0.5 % per year on continuous service. Annual zero-check and three-year recalibration recover catalog accuracy.
Realistic installed accuracy on a well-applied thermal mass meter: ±1.5 to ±2.5 % of rate over a 12-month interval. Coriolis on the same duty holds ±0.5 % through the same year. The 1.5× to 5× accuracy gap is what the 5 to 10× cost differential buys.
Final Thoughts
Picking between thermal mass and Coriolis is rarely about which meter is “better.” It is about matching the meter to the duty: the gas composition stability, the contract accuracy, the line pressure, and the install constraint that the existing plant dictates.
- Use thermal mass for compressed air, single-component gas (N₂, O₂, CH₄), variable-load duty with ±1 % accuracy budget, and any service below 0.3 MPa absolute where Coriolis loses signal.
- Use Coriolis for liquid duty, custody-billing accuracy tighter than ±0.5 %, multi-component or drifting gas streams, and any duty where mass + density + temperature on one instrument simplifies the install.
The 5 to 10× cost gap between thermal mass and Coriolis is real, but so is the 1.5 to 5× accuracy gap and the install-flexibility gap. Mastering both lets the buyer specify the right one without over-paying or under-spec’ing.
Not sure which is right for your duty?
Send service data (pressure, flow range, gas composition, line size, accuracy target) and we will return a sized quote within 4 business hours.