Vortex Flowmeter Gas Custody Transfer: Approved or Not

Start Here: No Gas Custody Standard Backs Vortex

Vortex shedding is not a recognized fiscal measurement technology for natural gas. The accepted lineage runs orifice, then turbine, then multipath ultrasonic, and each one has a measurement standard behind it (AGA Report No. 7 for turbine, AGA Report No. 9 for multipath ultrasonic) plus a flow computer that corrects to contract base conditions using AGA 8 compressibility and a live gas composition. Vortex has no equivalent gas custody document, so the counterparty, the regulator, or the auditor can reject the ticket no matter how well the meter behaves on the day.

Put vortex where it earns its keep: plant balance, allocation, fuel gas, check metering, flare header totals. Do not put it on the fiscal point and expect the number to survive a dispute.

Read the Panel Before You Blame the Meter

Most custody disputes are piping problems wearing a meter costume. Match the symptom first.

The Mechanism: These Are All Inferential Meters

None of these devices measure standard volume or mass directly. They measure a proxy and apply a coefficient that was established under a specific flow profile.

• Vortex: shedding frequency follows f = St * v / d, and the K-factor in pulses per unit volume is only linear above the manufacturer's minimum Reynolds number. Signal strength scales with the lift force on the bluff body, roughly proportional to density times velocity squared, so low-pressure gas gives a weak signal and turndown collapses. Read the minimum density and minimum Reynolds limits off the meter datasheet before anything else.
• Multipath ultrasonic: each chord returns a transit-time difference, and the meter weights those chord velocities into a mean pipe velocity. The weighting assumes a profile shape. Swirl, asymmetry, and a jet off a control valve break the assumption. That is how a meter certificated at 0.1 to 0.2 percent in a straight, dry, conditioned calibration lab lands 20 percent out in a real gas plant near bends and valves.
• Orifice: also inferential, but its discharge coefficient database is enormous, its installation rules are explicit in the standard, and its failure modes are visible when you pull the plate. That is why orifice is still the world's default fiscal gas meter.

Wet gas is a separate failure class. Liquid on the transducer ports, in the shedder wake, or across the orifice bore invalidates every one of these coefficients. Fix the gas conditioning before you argue about meter accuracy.

Procedure: Qualify the Meter Run

1. Get the contract and regulatory basis in writing. Find out which standard the fiscal point must meet and who audits it. That decision closes the technology question before you price hardware.
2. Fix the technology to a standard: orifice, turbine, or multipath ultrasonic. State the uncertainty target for the whole measurement system, not the meter alone.
3. Collect real process data: full composition including CO2 and inerts, water and hydrocarbon dewpoints, line pressure and temperature, minimum and maximum flow. If liquid drops out at line conditions, no inferential gas meter holds accuracy.
4. Measure the run as built. Record the distance in pipe diameters from the meter to every upstream and downstream bend, tee, reducer, thermowell, and control valve, then compare against the meter standard and the manufacturer's installation requirement.
5. Where straight length is short or a control valve sits close upstream, install a flow conditioner at the position the standard specifies. Do not assume a conditioner rescues a valve mounted a few diameters away.
6. Specify calibration at a high-pressure natural gas flow facility at line pressure and matching Reynolds number. Water or atmospheric air calibration does not transfer. Southwest Research Institute in the USA and NEL in Europe are the reference points auditors recognize.
7. Wire the gas chromatograph to the flow computer so composition updates automatically, and verify the compressibility method and base conditions the computer actually uses.
8. Design in verification: a check meter in series and continuous logging of meter diagnostics to the historian.

For hardware, the vendors with real custody transfer investment in this space are Daniel (Emerson), FMC Kongsberg, Instromet, and ABB. For flare gas and lower-cost duty where the accuracy target is looser, Roxar and Panametrics are worth quoting.

Verify Before You Sign the Ticket

• Compare the meter's measured speed of sound against the value computed from the GC composition, pressure, and temperature per AGA Report No. 10. Path-to-path spread outside the manufacturer's limit points to a transducer or a composition problem, not a flow problem.
• Log path velocity ratios, swirl and cross-flow indices, gain, SNR, and performance percentage at several flow rates. That fingerprint is your baseline; a healthy meter reproduces it.
• Run the series check meter for a full production day. A repeatable fixed offset is an installation effect. A drifting offset is fouling, liquid, or transducer degradation.
• On orifice runs, inspect the plate for upstream edge sharpness, bore diameter, flatness, and deposits, then verify DP zero at line pressure through the manifold.
• Audit the flow computer: base pressure and temperature, compressibility method, K-factor or discharge coefficient, and the parameter change log.
• Calibrate the static pressure and temperature transmitters. A one percent pressure error walks straight into the corrected volume.

Two fixes that waste time: swapping one vortex for a better vortex, and re-calibrating a meter whose diagnostics are already clean. Neither touches the flow profile that is causing the error.

Pitfalls That Cost Money on Gas Skids

• Treating datasheet accuracy as a field specification. It is a laboratory number obtained in ideal piping with clean dry gas.
• Reading parallel-meter disagreement as meter drift when the two runs have different upstream geometry.
• Installing ultrasonic meters in CO2-heavy or wet gas without checking attenuation margin and dropped-path behavior.
• Sizing a vortex for low-pressure gas, where density is too low to sustain shedding across the required range, then losing everything below the cutoff as unbilled gas.
• Mounting a vortex near reciprocating machinery and accepting the false counts as flow.
• Oversizing the meter so normal operation sits at the bottom of the linear range.
• Changing meter technology on an existing fiscal point without notifying the counterparty. That converts a metering issue into a contractual one.

Stop field work when the diagnostics sit inside the manufacturer's limits, the run meets the straight-length requirement in the applicable standard, and the meters still disagree beyond their combined uncertainty. At that point open a case with the meter manufacturer's measurement support group carrying the diagnostic logs, the as-built run drawing, and the gas analysis, and arrange re-proving at the flow facility that issued the calibration certificate. If money is already in dispute, bring the contract auditor in before anyone edits a flow computer parameter.

FAQ