Skip to content
Coriolis proving is a custody argument, not a sticker that says calibrated
All analysis

Sector · Plant systems · 11 Aug 2026 · 2 min

Coriolis proving is a custody argument, not a sticker that says calibrated

Master-meter runs, zero checks, and acceptance bands decide whether mass flow stays honest—plant measurement integrity, not generic calibration calendars alone, not historian tag quality essays, not soft-sensor model cards.

Coriolis meters sell direct mass and density. Plants buy them for batch honesty, material balance, and—sometimes—money. Then the sticker ages. A “calibrated” label from last turnaround does not prove the meter still agrees with a master under today’s fluid, temperature, and piping stress. Proving—master-meter comparisons, prover loops, or documented verification campaigns—is the industrial language that keeps Coriolis claims accountable.

Plant calibration management owns the broader metrology program. Historian tag quality owns units and flatlines. Soft-sensor cards own inferential gates. This brief owns Coriolis proving as a specific discipline: when the tube array is trusted enough to run the plant’s books.

What proving has to catch

| Drift source | What proving can reveal | | --- | --- | | Coating / erosion on tubes | Factor shift vs master | | Soft zero / installation stress | Bias that lab water never saw | | Electronics / config swaps | Factor tables that no longer match the sensor | | Fluid property extremes | Density path errors that “look fine” on water |

A meter that passes a water lab and fails on hot hydrocarbon is not a mystery. It is a duty mismatch you should have proved on the fluid that matters—or accepted with eyes open.

Coriolis mass flow meter on process piping

Nameplate accuracy is a claim. Proving is evidence.

Master-meter proving piping and valves

If the master is overdue, you only proved folklore.

A proving campaign that survives audit

  1. Freeze fluid, temperature band, and flow points in the procedure—not “as found.”
  2. Record zeros before and after; treat soft-zero habits as MoC events.
  3. Compare against a master with unbroken traceability—or a documented in-situ verification method the site owns.
  4. Set remove-from-service bands before the run, not after the argument.
  5. Tie results to MES/batch genealogy when the meter gates release.

An anonymized specialty-chemicals site that skipped proving for two years “because Coriolis is accurate” found a 1.8% factor shift after a coating campaign—enough to invent yield losses that production blamed on reactors. The reactors were fine. The books were not.

Adjacent fences

Calibration management owns the plant-wide program. Invoice autopsies own utility bill lines. Soft sensors own inferential quality. This page owns Coriolis proving as mass-flow truth. Do not tattoo “direct mass” on a P&ID and skip the master.

Share

LinkedIn

More in this sector

Plant systems