Direct mass-flow measurement does not guarantee single-phase reference accuracy when gas is entrained in a liquid. If a thermal-oil circuit shows fluctuating flow, changing density or abnormal diagnostics, establish the measurement state and the actual sensor, transmitter and software configuration before reaching a conclusion. A trend alone neither proves meter damage nor identifies a fluid-quality problem.
This guide supports procurement and instrumentation handover. It sets out documentation checks, not instrument settings, venting or modification procedures. It assigns no universal accuracy to any meter in two-phase thermal-oil service.
Retain the model, fluid, reference conditions, flow range and options attached to an accuracy statement. A meter described as suitable for liquids and gases is not thereby guaranteed to meet either single-phase specification for an arbitrary mixture.
Endress+Hauser explains that entrained gas changes the measuring tube's response and can introduce different, opposing error mechanisms.[1] Once gas is present, obtain applicable performance evidence rather than copying the single-phase specification into an acceptance statement.
Coriolis meters use the response of vibrating tubes to measure mass flow and can use resonance characteristics to measure density. They do not first multiply volume flow by an assumed fixed liquid density to obtain mass flow.[1] Direct measurement, however, does not eliminate second-phase effects.
The manufacturer's white paper states that, with entrained gas, the reported mass includes both gas and liquid; density and subsequently volume flow can also be affected.[2] This identifies the measurement object, not an error-free result. Specify whether the task concerns total mixture mass, net liquid mass, operating volume or reference-state volume. Without additional information and an applicable model, the display cannot be treated as a gas-corrected net liquid quantity.
For unit and state-basis questions, see comparing thermal-oil mass and volume flow. Aligning units does not remove two-phase measurement error.
| Mechanism | Behavior described by the manufacturer | Interpretation limit |
|---|---|---|
| Bubble effect associated with free bubbles | Bubbles do not follow the surrounding liquid or tube motion exactly; mass flow and density generally underread | Not a universal direction for every gas-containing liquid |
| Resonator effect associated with suspended microbubbles | Changes in mixture acoustics interact with tube vibration and can cause mass flow and density to overread | A single discrepancy cannot identify the bubble state or justify a fixed correction |
These mechanisms are described in the cited manufacturer sources.[1][2] Bubble size and distribution, viscosity, pressure, flow regime and instrument configuration affect the response. When bubble types coexist, the combined indication may change with conditions. Neither “gas always causes under-registration” nor its opposite is a safe general conclusion.
A gas fraction also needs its pressure, temperature, definition and measurement location. Dispersed microbubbles, large pockets and slug flow are different conditions. This article establishes no universal allowable gas percentage.
Gas may be associated with ingress, release from solution or other process changes. A correlation with liquid level or pressure is a lead, not proof of origin. Distinguish entrained gas from a second phase formed by vaporization: not every bubble is air, and vaporization or cavitation cannot be diagnosed without fluid, pressure and equipment evidence.
Align flow, density, temperature, pressure, level, diagnostics and operating events. Check real flow changes as well as signal settings and transmission. The related guide on thermal-oil pump cavitation evidence addresses a different question; meter disturbance and pump cavitation are not one automatically confirmed cause.
Drive, damping, signal-quality and medium-inhomogeneity information can help identify a changing measurement state, but definitions are model-specific. Relevant diagnostic indices in the cited sources describe relative disturbance, not a one-to-one gas volume fraction.[1][2] Do not add a percent sign to an index and call it a measured gas fraction. A dedicated quantitative output is a separate capability requiring configuration and definition checks.
The official Gas Fraction Handler (GFH) explanation explicitly distinguishes signal stabilization from error correction: smoothing and diagnostics do not themselves correct measurement error.[1] A steadier display or improved repeatability does not prove that single-phase accuracy has returned. Review the effect of processing on dynamic response, short transfers and control duties.
Multi-Frequency Technology uses different vibration modes to provide additional information and compensate for the resonator effect.[1][2] It is not ordinary filtering or a default feature of every Coriolis meter. Nor does it guarantee one error limit for every liquid, gas fraction and flow regime. Tests on other fluids cannot simply become a thermal-oil contractual guarantee.
The sources discuss reducing gas entry, appropriate pressure, sizing and installation as application-design considerations.[1][2] An actual assessment must also consider design pressure, pump duty, pressure loss, vaporization risk, temperature and model-specific requirements. General guidance is not an instruction to adjust a valve, raise pressure or reduce meter size on site.
Do not discharge hot fluid, disconnect equipment or bypass protection to investigate gas effects. Authorized personnel must assess modifications and verification under project and site procedures. No universal orientation approval, backpressure setting or sizing value is prescribed here.
This is an editorial evidence-organizing tool, not a manufacturer's inspection procedure. Add the source, revision, responsible person and open issue for each row. Mark missing evidence as pending.
| Check | Evidence to retain |
|---|---|
| Measurement objective | Total mass, net liquid quantity, density or volume and its state basis |
| Instrument identity | Sensor/transmitter models, serial numbers, software and options |
| Process and phase state | Fluid, temperature, pressure, flow, second-phase condition and supporting evidence |
| Accuracy applicability | Single- or two-phase statement, coverage, limits and project acceptance requirements |
| Diagnostic meaning | Variable definitions and synchronized records; separate indices from quantitative gas fraction |
| Function status | Actual filtering, GFH or compensation configuration, enabled state and applicability |
| Comparison and acceptance | Reference method, time window, starting/ending inventory, error limit and criterion |
| Closure | Verified facts, possible explanations, verification records and unresolved questions |
If a totalizer differs from a reference, first establish the reference method, coverage and time window. Weighing comparisons need to account for pipe and vessel inventory and transfer start/end conditions. Volume comparisons also need consistent temperature, density and reference-state definitions. Then align diagnostics with process events; one totalized discrepancy is not a permanent correction factor.
Continuous output, no fault alarm, a smooth display and an acceptable totalized result are different facts. A contract for single-phase liquid performance needs evidence of phase conditions. A requirement for gas-containing liquid measurement needs separately agreed conditions and performance evidence.
Not necessarily. Calibration evaluates performance under specified conditions. Gas-related effects can vary with the process and act in different directions. Establish the process state and measurement mechanism first; do not use a zero adjustment or fixed factor to conceal an unexplained two-phase problem.[1]
A Coriolis meter cannot unconditionally retain its single-phase accuracy specification when thermal oil contains entrained gas. Define the quantity and phase state, distinguish bubble mechanisms, and assess diagnostics, stabilization and compensation separately. The final judgment needs the actual model, operating conditions and traceable verification—not a smooth trend alone.
[1] Endress+Hauser, Coriolis flowmeter accuracy with entrained gas. Official explanations of bubble and resonator effects, GFH, diagnostic indices and multi-frequency capability, reviewed October 30, 2026. Numerical examples for other fluids are not adopted as thermal-oil guarantees.
[2] Endress+Hauser Flowtec AG, Paul Ceglia and Hai-Thuy Ngo, Entrained gas handling in Promass Coriolis flowmeters, May 2020, 8 pages. Reviewed October 30, 2026: page 2 on measurement object and bubble types; pages 3–5 on application design, MFT and GFH; pages 6–7 on diagnostic indices. Configuration descriptions from 2020 do not establish default capability for every current model.
No universal gas limit, error coefficient, field procedure or model-specific thermal-oil approval is given. A measurement abnormality is not automatically evidence of defective heat transfer fluid.
