Flow-Measurement Technologies & Installation Effects

Flow is motion through a defined surface. Volumetric flow rate is

$$Q=\int_A \mathbf{v}\cdot d\mathbf{A},$$

and mass flow rate is

$$\dot m=\int_A \rho\mathbf{v}\cdot d\mathbf{A}\approx\rho Q.$$

The approximation assumes a representative density. Gas “standard volume” is a mass-derived quantity converted to a declared reference temperature, pressure, and compressibility convention; “Nm³/h” without that basis is incomplete.

Select the measured interaction #

The Endress+Hauser technology overview groups several current methods. Each observes a different physical effect:

No technology is universally “most accurate.” Accuracy, turndown, zero stability, pressure drop, response time, maintenance, and installed uncertainty depend on the application.

Core mechanisms #

For an obstruction meter, Bernoulli and continuity lead to an ideal square-root relation:

$$Q=C A_2\sqrt{\frac{2\Delta p}{\rho(1-\beta^4)}},$$

where $C$ incorporates discharge behavior and $\beta$ is diameter ratio. Real use needs applicable standards, tap geometry, Reynolds correction, compressibility for gases, and primary-element tolerances.

An electromagnetic meter follows Faraday induction:

$$E\propto B D v,$$

with magnetic field $B$, electrode spacing related to diameter $D$, and average velocity $v$. Conductive liquid must contact the electrodes and fill the measuring section.

Vortex shedding follows approximately

$$f=\mathrm{St}\frac{v}{d},$$

where $d$ is bluff-body scale and Strouhal number $\mathrm{St}$ is nearly constant only over the characterized Reynolds range.

Coriolis meters drive tubes in vibration; moving mass creates phase or deformation related to mass flow, while resonant behavior can provide density. This directness does not make them immune to entrained gas, external vibration, or mounting stress.

Installation changes the measurand #

Elbows, valves, reducers, partially open dampers, pumps, swirl, pulsation, and asymmetric headers distort the velocity profile. A meter calibrated with a developed profile may bias when installed immediately after a disturbance. Straight-run requirements are technology- and disturbance-specific; flow conditioners can help but add pressure loss and need their own installation rules.

flowchart TB
  D["Upstream disturbance"] --> P["Velocity, phase, or acoustic profile"]
  P --> M["Meter interaction"]
  M --> S["Raw sensor signal"]
  S --> C["Compensation and output"]

The meter must be full when its method assumes a full pipe. Avoid gas pockets in liquid service and liquid accumulation in gas service. Respect required orientation, flow direction, grounding, electrode position, transducer acoustic path, and available backpressure. Locate control valves according to cavitation, flashing, and meter requirements rather than a generic upstream/downstream rule.

Two-phase flow is not a small disturbance. Bubbles change density and acoustic paths; droplets change thermal and vortex behavior; slip means phases move at different velocities. If two-phase operation is expected, choose a method and model explicitly qualified for it.

Range and dynamics #

Size by actual minimum, normal, maximum, startup, upset, and reverse flow—not pipe nominal diameter alone. Oversizing reduces velocity and usable signal; undersizing increases pressure loss, erosion, cavitation risk, and peak velocity.

Turndown is the ratio of maximum to minimum usable flow under declared performance. At low flow, zero offset, noise, buoyancy, leaks, and cutoff logic can dominate. A low-flow cutoff stabilizes totalization but intentionally suppresses real flow below its threshold; document it.

Pulsating flow requires sufficient sensor and output bandwidth plus time-aligned density/pressure/temperature compensation. Damping can calm a display while biasing totals or hiding control-loop dynamics.

Uncertainty and totalization #

Propagate every variable used in the measurement equation: primary sensor, dimensions, density, pressure, temperature, composition, installation correction, analog/digital conversion, and time base. A percentage-of-rate specification behaves differently from percentage-of-span near zero.

Totalized quantity is the time integral

$$V=\int_{t_0}^{t_1}Q(t)\,dt.$$

Clock error, sampling, dropout, reverse-flow policy, reset behavior, and gap filling become measurement terms. Verify both instantaneous output and totalizer behavior.

Commission and diagnose #

1. Confirm fluid, phase, composition, range, pressure, temperature, and hazardous-area data.
2. Inspect pipe size, liner/gasket intrusion, orientation, supports, grounding, straight runs, and valve/pump locations.
3. Zero only under the manufacturer's required no-flow and full-pipe conditions.
4. Verify engineering units, density/reference conditions, direction, damping, cutoff, pulse scaling, and totalizer rollover.
5. Compare mass balance, portable reference, or process invariants across operating points.

If error changes with valve position or flow rate, suspect profile, Reynolds regime, cavitation, or compensation. If zero drifts while isolated, suspect bubbles, coating, mechanical stress, vibration, or electronic offset. If two meters disagree, first reconcile their measurands and reference conditions before declaring either faulty.