For homogeneous liquids in a fully charged pipe, transit-time ultrasonic flowmeters deliver ±1.0% of reading accuracy with 400:1 turndown and bi-directional ranges of 0.01–12 m/s, making them the default pick over magnetic meters when the fluid is non-conductive (oils, alcohols, deionized water, hydrocarbons) [S2][S4].
Selection starts with the measurement principle (transit-time vs Doppler), the transducer form (clamp-on, insertion, or pipe-section), and the operating envelope (DN, fluid temperature, particulate load), all of which interact with accuracy and cost more than vendor brand does [S1][S4].
Transit-time vs Doppler: which principle matches the fluid
Transit-time meters measure the difference between upstream and downstream sound propagation through a clean liquid; Doppler meters rely on frequency shift from particles or bubbles, so they need entrained solids or gas to operate and cannot be used on ultra-pure water or clean chemicals [S2].
For clean water, potable water, sea water, hot water, cooling water, thermal oils, alcohols, acids, and chlorine, specify transit-time; the SCL-80/81/82 product family in [S1] is representative, with a stated fluid envelope of water, wastewater, and other homogeneous liquids in a fully charged conduit, suspended solids limited to ≤10 g/L [S1].
Doppler remains useful for raw wastewater, slurries, and partially filled flows where bubbles or particles are guaranteed, but its accuracy is materially lower and its zero stability is poor; treat it as a dedicated dirty-application tool, not a general substitute [S2].
Clamp-on, insertion, and inline: form factor drives install cost and accuracy
Clamp-on transducers mount externally with no pipe cut, no pressure drop, and zero process contamination risk, and they cover the widest diameter range: DN15 to DN6000 across commercial product lines [S2][S4]. The Jade Ant series quotes ±1.0% accuracy, -40°C to +200°C liquid temperature, and flow velocity 0.01–25 m/s, which is the realistic ceiling for non-intrusive transit-time work [S4].
Insertion probes (SCL-80 in [S1]) require a hot-tap or pipe-section mounting and trade installation effort for better signal coupling and lower sensitivity to air bubbles or scale on the pipe wall; they are the practical step up when the pipe is old, lined, or has thick walls that attenuate clamp-on signals [S1].
Pipe-section / inline meters (SCL-82 in [S1]) bolt between flanges and behave like a spool piece; they are the right call when permanent custody-grade accuracy is required, when the line is small (under DN50 where clamp-on accuracy degrades), or when the fluid is dirty and a defined acoustic path improves repeatability [S1]. For a deeper look at the in-pipe alternative, see the related guide on vortex flowmeter sizing and selection traps.
Sizing math: Reynolds, velocity window, and turndown
Engineers consistently get ultrasonic sizing wrong by skipping the Reynolds check and the low-velocity cut-off. Below roughly 0.3 m/s, the stated accuracy figure must be re-read against the absolute spec, not the percentage spec, because percentage-based claims collapse near zero [S2].
The Lanry TF1100-EC datasheet makes this explicit: ±1.0% of reading at velocities above 0.3 m/s, ±0.003 m/s absolute below 0.3 m/s, with 0.25 mm/s resolution, 0.2% of reading repeatability, and 0.5 s response time [S2]. Treat that 0.3 m/s threshold as a universal rule of thumb when comparing ultrasonic bids.
For pipe sizing, pick a normal operating velocity between 1.0 and 3.0 m/s to stay inside the high-accuracy band, then verify the minimum nighttime flow does not drop below 0.05–0.1 m/s; if it does, specify a smaller bore or a different technology (Coriolis or magnetic) for that leg, and reserve the ultrasonic for the main line, as covered in the 2026 Coriolis flowmeter supplier map.
Fluid, pipe, and environment limits you must verify
Ultrasonic meters demand a fully charged pipe: partial fill, air pockets, or heavy scale will absorb the signal before it reaches the receiver. The SCL series datasheet explicitly states "fully charged conduit" with suspended solids ≤10 g/L, which sets the upper bound for standard transit-time units [S1].
Material compatibility is broad on the outside: steel, cast iron, cement, PE, and PVC are all listed as transducer-mountable pipe materials, but lined pipes, pipes with internal coating, or concrete pipes with steel liners require a site test before commitment [S1]. Liquid temperature ranges from -40°C to +200°C across the commercial product lines surveyed, with -35°C to 200°C quoted by the Lanry TF1100-EC and -40°C to +200°C by the Jade Ant wall-mounted unit [S2][S4].
Power and signal: typical wall-mounted units accept isolated DC 8–36 V or AC 85–264 V, draw 50 mA operating current, and offer 4–20 mA analog outputs; portable or remote units ship with a built-in high-capacity battery for sites without power [S4]. Always confirm output count (3-channel 4–20 mA in, plus pulsed and communication outputs) against the DCS card budget before ordering [S4].
Application matrix: which variant fits which duty
Clamp-on is correct for retrofit on existing steel or iron lines, for temporary measurement during commissioning, and for any non-conductive fluid where an electromagnetic meter cannot work (oils, alcohols, deionized water, hydrocarbons) [S4]. It is the wrong pick for custody transfer, for dirty wastewater above 10 g/L solids, or for pipes below DN50 where transducer geometry starts to dominate the error budget [S1][S4].
Insertion is correct for medium-to-large lines (DN100 and up) in chemical, petrochemical, and power plants, where the line can be hot-tapped under pressure, where the fluid is hot (steam condensate, thermal oil), and where a permanent but lower-cost installation beats a full spool-piece [S1][S2].
Inline / pipe-section is correct for new builds, for small-bore clean-service lines, and for custody or balance-of-plant measurement where traceability matters; expect 3–5x the installed cost of a clamp-on but gain in accuracy stability and long-term zero drift [S1]. Mining and slurry service is a different game, and the practical spec path is laid out in the Best ultrasonic flowmeter for mining guide.
Decision criteria: clamp-on, insertion, inline side by side
Across the three form factors, the trade-off is consistent: clamp-on wins on install cost and zero process interruption, insertion wins on signal quality and temperature tolerance, and inline wins on accuracy stability and traceability; cost rises in the same order, typically with a 2–3x step from clamp-on to insertion and another 2–3x step to inline at the same diameter [S1][S2][S4].
For diameter range, all three cover DN50–DN600 with the standard transducer set, but clamp-on extends to DN6000 with low-frequency pairs while insertion is rarely offered below DN80; inline is available down to DN15 but at materially higher unit cost per diameter [S2][S4]. Maintenance follows the same pattern: clamp-on transducers are field-replaceable in minutes, insertion probes need a hot-tap retraction tool, and inline units require a line shutdown and spool replacement, so plan your construction machinery and equipment access and isolation valves at the design stage.
Sourcing, standards, and what to verify on the datasheet
For trade buyers sourcing on Alibaba-style platforms, confirm the manufacturer publishes the absolute low-velocity accuracy (e.g. ±0.003 m/s below 0.3 m/s), not just a percentage number, because the percentage figure misleads near zero [S2]. Ask for the repeatability figure as a percent of reading, the response time in seconds, and the full operating temperature envelope, then check the power supply range covers the worst-case plant condition (AC 85–264 V covers most global builds) [S2][S4].
Standards to verify by application: ISO 4064 for clean-water custody, OIML R49 for water meters, API MPMS chapters for hydrocarbon service, and ATEX/IECEx for hazardous-area zones; the meter body and cable glands both need to match the zone classification, and the cable gland sizing math is independent and covered in the cable gland sizing guide. For a general flow-instrument primer, the ultrasonic flowmeter encyclopedia page sets the baseline terms.
Next trackable signals: ask each bidder for the calibration certificate with traceability to a national institute (NIST, NIM, NPL), the wet-cal vs dry-cal method, and the upstream/downstream straight-pipe requirement (typically 10D upstream and 5D downstream for transit-time), because a meter that is technically ±1.0% on the bench can easily deliver ±5.0% in the field if the piping violates the run requirement.
Spec-level background on the components involved: linear guide.