Selection of an ultrasonic flowmeter is dominated by four engineering parameters — fluid sound velocity, particulate or gas content, available straight-pipe length, and pipe material access — and the SGM LEKTRA product entry on DirectIndustry (2026-05-17) is a concrete data point: the listing explicitly classifies a modern ultrasonic meter as "Technology ultrasonic, Venturi / Type volume / Installation threaded," meaning the same instrument family can be built into a Venturi spool or a threaded inline body depending on the process header [S1].
The application envelope stated by manufacturer guidance is unusually wide for a single flow technology: "oil industry, metallurgical industry, petrochemical industry, power industry" plus water, gas and a broad medium set, and the installation is "directly on the outer wall of the pipelines, not affected by the dielectric effect, easy installation fast, without cutting process piping" [S4]. That breadth is misleading on its own — the same guidance restricts the technology to situations where the medium's ultrasonic transmission is acceptable, which is the actual selection gate.
Two sensing principles, two different fluid windows
Transit-time ultrasonic flowmeters require a clean, homogeneous liquid with relatively stable sound velocity; Doppler ultrasonic flowmeters require a defined minimum concentration of scatterers (bubbles or particles) to function, and are the only ultrasonic variant that can handle slurries, raw wastewater, and two-phase flows with visible aeration [S4]. The Free Dictionary entry defines the instrument family as "an instrument that measures the rate of flow of a liquid or gas within a pipe or tube" — a deliberately broad definition that the specifier must narrow with the principle-of-operation choice [S2].
Practical implication: if the line is demineralized water, condensate, refined hydrocarbon, or any other low-scatter fluid, transit-time is the only working principle; if the line is raw sewage, dredge slurry, or pulp stock, Doppler is the only working principle. Specifying a Doppler unit on a clean line, or a transit-time unit on an aerated line, is the single most common ultrasonic-flowmeter field failure — the meter reads a value, the value is wrong, and the diagnostic on the HART layer usually shows a low signal-to-noise ratio, not an outright fault.
Clamp-on vs inline: the access decision
Clamp-on transducers mount on the outside of existing pipework, do not cut the process, and are "not affected by the dielectric effect" of the medium — they respond to the mechanical/elastic properties of the fluid, not its electrical conductivity [S4]. Inline (wetted) ultrasonic meters, of which the SGM LEKTRA threaded/Venturi type is one example, integrate the transducers into a flow body and are specified where the application needs factory calibration, permanent installation, or measurement of fluids where clamp-on acoustic coupling is marginal [S1].
For retrofits on existing carbon-steel or stainless headers where hot-tap is unacceptable and the pipe is accessible, clamp-on is the default. For new builds, custody-transfer class measurement, high-pressure steam condensate, or any service above the temperature ceiling of the clamp-on couplant, inline wetted is the default. Spire Metering's product line illustrates the third commercial branch — utility-grade ultrasonic meters with AMR/analytics bundles aimed at district heating and water distribution, where the deliverable is a metered data stream rather than a process-control 4-20 mA loop [S6].
Fluid and pipe constraints that decide the result
Manufacturer guidance is explicit on the physical limits: ultrasonic flowmeters "can be used to measure flow and provide water data" but the medium has to be acoustically conductive enough for the signal to cross the pipe wall, and the pipe has to admit a usable straight run upstream and downstream of the transducers [S3]. The same SL1188E-class unit bundles flow, pressure, turbidity, residual chlorine and pH into a single GPRS-linked node, which signals where the technology is being positioned in 2026: as a multi-parameter district-metering element rather than a single-variable flow device [S3].
Selection gates that have to be checked before any vendor shortlist:
1. Fluid: clean liquid (transit-time) vs particulate / aerated (Doppler). Gas and steam are possible with special high-frequency transit-time designs but are a separate specification exercise.
2. Pipe: material (carbon steel, stainless, ductile iron, PVC, lined pipe), wall thickness, and whether a clamp-on transducer can be coupled to it. Lined or plastic pipes need velocity-of-sound correction; some liners absorb ultrasound to the point that clamp-on becomes impractical.
3. Straight run: most inline transit-time meters need 10D upstream and 5D downstream minimum, and that figure is conservative for valves, pumps and reducers immediately upstream.
4. Temperature/pressure: clamp-on couplant limits roughly -40 to +200 °C depending on the gel; inline wetted meters are specified per flange class and process temperature.
Comparison against the other flow technologies
For clean conductive liquids, an electromagnetic flowmeter is usually cheaper and more accurate than an inline ultrasonic; for clean liquids where conductivity is variable or non-conductive (hydrocarbons, deionized water), ultrasonic is the only non-intrusive option. For high-accuracy mass flow or for any process where density varies, a Coriolis flowmeter replaces both, at a higher unit cost and with a larger installed footprint — the Coriolis Flowmeter Buying Guide 2026 walks through that trade-off in detail. For clean, low-viscosity, mid-to-high flow rates in custody transfer, turbine and positive-displacement meters still dominate on cost per accuracy point. [S1]
Criteria-based comparison (clean liquid service, 2 in to 12 in line, ANSI 150#):
— Ultrasonic clamp-on: medium accuracy (±1-2 % of reading typical), no pressure drop, no pipe cut, multi-parameter option [S3]; fails on dirty, aerated, or lined pipe.
— Ultrasonic inline: better accuracy (±0.5-1 %), factory-calibrated body, handles higher T/P [S1]; requires process shutdown to install and a wetted material spec.
— Electromagnetic: ±0.3-0.5 %, no moving parts, conductive liquid only; price-competitive at smaller line sizes [S2].
— Coriolis: ±0.1 %, direct mass flow and density, multiphase-capable; high unit cost, significant pressure drop [see Coriolis guide].
Who ultrasonic is for — and who it is not for
Specifying ultrasonic pays off when at least three of these hold: the fluid is acoustically clean, the pipe is accessible and cannot be cut, the operator wants a non-intrusive installation with no pressure drop, and the measurement is for monitoring or district metering rather than fiscal metering of hydrocarbons. The 2026 commercial evidence supports this positioning: utility metering with built-in GPRS data push [S3], retrofit clamp-on probes for process lines where the operator refuses a hot tap [S4], and inline Venturi / threaded bodies for OEM skid integration [S1].
It does not pay off for: high-accuracy fiscal measurement on hydrocarbons (Coriolis or turbine still own that segment), two-phase or strongly aerated streams without enough scatterers for Doppler, lined or thick-walled plastic pipes that attenuate the acoustic beam below a usable SNR, or any service where the line is so short that the 10D/5D straight-run requirement cannot be met without an expensive spool piece. The Dalian Sonic Meter product range is a useful indicator of the technology's commercial centre of gravity: ultrasonic flowmeters, ultrasonic heat meters, and differential-pressure flowmeters, all of them aimed at district heating, water distribution and light-industrial flow — not at upstream oil & gas custody [S5].
Installation and commissioning watch-list
Four field conditions decide whether a 2026 ultrasonic installation reads correctly on day one: [S2]
— Couplant: clamp-on units depend on a continuous acoustic-coupling layer (gel, grease or epoxy pad). Any air gap drops the signal below the detection floor.
— Pipe wall paint, scale and liner: scale, internal liner thickness, and external paint thickness all shift the apparent path length and velocity, and must be entered as commissioning parameters — not guessed.
— Upstream disturbances: a single partially-closed valve 5D upstream will distort the velocity profile; the meter still produces a number, and the number is wrong.
— Temperature transients on outdoor clamp-on installations: couplant viscosity changes with temperature, and on district-heating return lines the transit-time amplitude can swing 20-30 % over a 24-hour cycle. A diagnostic channel (signal strength, SNR, flow profile factor) is the right acceptance criterion, not a single point reading.
Standards, sourcing and where to verify
The relevant cross-industry references for ultrasonic custody and non-custody use include ISO 4064 (water meters), EN 1434 (heat meters), AGA Report No. 9 (ultrasonic gas), and the IEC 60079 family for hazardous-area electronics. For hazardous-area installation of an ultrasonic flowmeter, the Pt100 RTD selection guide shows how the same IEC 60079 logic carries across instrumentation on a skid, and the differential pressure transmitter buying guide covers the comparator case where ultrasonic and DP compete head-to-head. The technology's working definition — an instrument that measures the rate of flow of a liquid or gas within a pipe or tube [S2] — has not changed; what has changed since 2024 is the bundling of ultrasonic flow with pressure, turbidity, residual chlorine and pH on a single GPRS-linked node [S3], which makes the 2026 selection question less "which ultrasonic meter" and more "which multi-parameter ultrasonic platform fits the data architecture of the plant."