A transit-time ultrasonic meter measures flow by timing acoustic pulses sent across the pipe, so it depends on those pulses actually arriving at the far transducer with enough strength to be timed. When something between the transducers absorbs, scatters, or blocks the sound, the meter stops receiving a usable signal on that path, and the path drops out. A multi-path meter can lose one chord and keep running on the rest, which is a strength, but it also means the flow number can quietly degrade or freeze while looking fine. This page explains why paths drop out, how per-path diagnostics reveal it, how a multi-path meter degrades gracefully, and why SCADA should alarm on path health rather than the flow value alone.
Ultrasonic Signal Dropout in one line: Ultrasonic flow meter signal dropout is the loss of a usable acoustic signal on one or more measurement paths, so the meter can no longer time pulses across those paths. It is caused by anything that absorbs or scatters the sound between transducers, including gas bubbles, solids or slurry, highly attenuating liquid, fouled or coated transducers, and an empty or partially filled pipe. When a path drops out, a multi-path meter can keep measuring on its remaining paths at reduced accuracy, or the meter may hold its last-good value, which is why monitoring per-path signal health matters more than watching the flow number alone.
A transit-time ultrasonic meter sends acoustic pulses diagonally across the flowing fluid between pairs of transducers and measures the tiny difference in travel time with and against the flow to compute velocity. For that to work, each pulse has to cross the pipe and reach the opposite transducer with enough amplitude to be detected and timed accurately. Anything that weakens or disrupts the sound on the way across threatens the measurement, and when the received signal falls below what the meter can reliably use, that path drops out.
The most common culprit is gas bubbles or entrained gas, because the sharp density change at each bubble surface scatters and reflects the acoustic energy, so much of the pulse never reaches the far side. Solids or slurry do the same by scattering and absorbing the sound, and a highly attenuating liquid, one that naturally absorbs acoustic energy, can weaken the signal to the point that longer paths no longer make it across. On the equipment side, fouled or coated transducers lose their ability to launch and receive sound efficiently, as buildup on the transducer face muffles the signal from both directions. And an empty or only partially filled pipe removes the medium entirely along some or all of a path, so there is nothing to carry the sound.
The signature of all these is the same at the meter: the received signal on the affected path weakens, the meter increases its receiver gain to try to compensate, and if the signal cannot be recovered the path is declared failed. Some conditions are transient, a slug of gas passes and the path recovers, while others are persistent, such as a fouled transducer or a chronically aerated stream, and keep the path down until the underlying cause is addressed. Either way, the flow measurement on that path is either lost or running on a marginal, gain-boosted signal that is more susceptible to error.
Modern ultrasonic meters expose a set of diagnostics per path that tell you the health of each chord rather than just the aggregate flow. Signal-to-noise ratio indicates how strong the received pulse is relative to background noise, and a falling SNR is an early sign a path is heading toward dropout. Receiver gain shows how hard the meter is working to detect the signal, and a gain that climbs and eventually saturates indicates the meter is straining to hear a weakening pulse before it gives up. Additional indicators such as signal quality, performance, or accepted-pulse counts round out the picture of whether each path is healthy, marginal, or failed.
The value of a multi-path meter is that it can lose a chord and keep measuring, degrading gracefully instead of failing outright. With several paths spanning the pipe cross-section, the meter combines them into a velocity profile, and if one path drops out, it can continue on the remaining paths, often substituting or reweighting to maintain an estimate. This resilience is genuinely useful, a brief gas slug that knocks out a path need not blank the whole measurement. But graceful degradation has a hidden cost: running on fewer paths means the meter is sampling less of the flow profile, so its accuracy declines even though it still produces a number.
That is the trap operators fall into. A four-path meter running on two paths still shows a flow value, and nothing on a basic display shouts that half its measurement capability is gone. The flow number looks normal, so it gets trusted at its full stated accuracy when in reality it has degraded. The per-path diagnostics are exactly what reveal this, they distinguish a healthy meter reading four good paths from a compromised one limping on two, but only if someone is watching them. The flow number alone cannot tell you how it was produced.
The right way to trust an ultrasonic meter is to alarm on path health and signal diagnostics, not just on the flow value. If a system only watches flow, it stays silent while the meter quietly falls back onto fewer chords and its accuracy erodes, and it only reacts when the number itself does something obviously wrong. By monitoring per-path SNR, gain, and path status, you catch the degradation while the flow reading still looks plausible, which is precisely when you want to know, before a marginal measurement is used for custody or control decisions.
This is a strong argument for pulling the meter's full diagnostic set into the monitoring layer rather than reading only its primary output. When a platform like Merobix collects per-path signal-to-noise, gain, and chord status over time, it can alarm the moment a path drops out or a gain starts saturating, and it can flag that the meter is now running on fewer paths than its accuracy specification assumes. Instead of trusting a flow number that is quietly running on two of four chords, the operator sees explicitly that the meter is degraded and can weigh its reading accordingly or dispatch someone to clear the cause.
Trending these diagnostics also turns intermittent, hard-to-catch problems into visible patterns. A path that drops out for a few minutes whenever the process aerates, a gain that creeps up over weeks as a transducer fouls, or a chord that fails every time the pipe runs partially full all show up as recurring signatures in the history rather than as fleeting glitches nobody was watching. That history distinguishes a transient upset from a developing fault, lets maintenance target the actual cause, whether it is transducer cleaning, an aeration source, or a piping fill problem, and preserves a record of exactly when the meter's accuracy was compromised so any affected measurement can be reviewed.
It loses signal when something between the transducers weakens or scatters the acoustic pulse so it cannot be reliably timed. Gas bubbles, solids or slurry, and highly attenuating liquids all scatter or absorb the sound, while fouled or coated transducers lose their ability to launch and receive it, and an empty or partially filled pipe removes the medium entirely. When the received signal drops below what the meter can use, the affected path drops out.
Yes, a multi-path meter can lose a chord and keep measuring on its remaining paths, which is a deliberate design strength that lets it survive a brief gas slug without blanking the whole reading. The catch is that running on fewer paths means it samples less of the flow profile, so its accuracy declines even though it still produces a number. That degraded number can look normal on a basic display, which is why watching path status matters.
The key per-path diagnostics are signal-to-noise ratio, which shows how strong the received pulse is, and receiver gain, which shows how hard the meter is working to detect it. A falling SNR and a rising or saturating gain both indicate a path heading toward dropout. Path status, signal quality, and accepted-pulse counts round out the picture of whether each chord is healthy, marginal, or failed, which is far more informative than the flow number alone.
Merobix reads your field devices into a cloud SCADA - the real thing behind these terms, live in days from any browser.