A clamp-on meter installed in bad hydraulics does not throw an error. It reports a confident, precise, wrong number. Four things decide whether your meter reads correctly, and only one of them is the meter.
Ask around a few plants and you will find someone who will tell you clamp-on ultrasonic flow meters do not work. They bought one, it read nonsense, and it went in a cupboard.
Almost every time, the instrument was fine. The installation was not. And the failure was invisible, because a clamp-on meter installed in bad hydraulics does not throw an error — it reports a confident, precise, wrong number.
Four things decide whether your meter reads correctly. Only one of them is the meter.
The meter measures velocity along an acoustic path and infers the mean velocity of the whole pipe from it. That inference is only valid if the flow profile is fully developed and symmetric.
An elbow, a valve, a tee, a reducer, or a pump destroys that. The profile becomes skewed, or it develops swirl, or both. The meter faithfully reports the velocity on its acoustic path — which is no longer representative of the pipe. The number it gives you is real. It is just not the number you wanted.
An elbow produces a skewed profile, and a skewed profile re-develops fairly quickly. A pump produces swirl — the whole body of fluid is rotating as it moves down the pipe. Swirl is stubborn. It decays over a much longer distance, and while it persists, the velocity vector along your acoustic path is not aligned with the pipe axis. That is a hard error, and it does not look like one.
Install in bad hydraulics, get a plausible number, and let it into a report or a control system as if it were good. Everyone downstream of that number will trust it, and nobody will remember the elbow.
The signal has to get through the pipe wall, twice. Some pipes cooperate.
Carbon steel, stainless steel, copper, PVC, HDPE. Thin wall is easier than thick wall. Smooth bore is easier than rough.
Cement-mortar-lined ductile iron. The classic. If the mortar has delaminated from the pipe wall, there is an air gap inside the pipe. You cannot reach it and you cannot fill it, and air stops ultrasound dead. Test before you commit to this pipe.
Heavy internal scale or corrosion. Attenuates the signal and — worse — changes the real internal diameter, which corrupts the area term in Q = v × A.
Coarse cast iron. The graphite structure scatters ultrasound.
Concrete. Generally defeats transit-time. Doppler sometimes copes — the Compu-Flow C6 lists concrete among supported materials.
Composite, GRP, fiber-reinforced. Case by case. Test.
Any air gap anywhere in the acoustic path is an acoustic wall. Behind a liner, under scale, in the couplant, between transducer and pipe. This one sentence explains nearly every no-signal condition you will ever meet.
This one is invisible and it is everywhere.
The instrument does not measure the pipe. It computes cross-sectional area from the dimensions you enter: outside diameter, wall thickness, liner thickness, liner material. Those numbers go straight into the flow calculation.
A wall thickness that is off by 10% puts a systematic error into every reading the meter will ever take, and the display will show it to four decimal places with complete confidence.
A thirty-year-old line with scale and corrosion is not the thickness the drawing says. Take an ultrasonic wall thickness measurement at the actual mounting location. The METRI Ultra ProLite has an integral wall thickness gauge (2–24 mm) built in precisely for this. It is the highest-leverage sixty seconds in the whole procedure.
Between the transducer face and the pipe there is a microscopically thin film of air. Ultrasound crossing solid→air→solid loses nearly all its energy at each boundary. A thin layer of air is, acoustically, a wall.
Couplant displaces it. That is its entire job.
A thin smear leaves gaps. Gaps are air. Air is a wall.
Standard gel on a hot line will run, then bake out. Your signal will fade over the course of the survey in a way that looks exactly like a real flow change — and on a permanent installation, it will fade over months and take the meter with it.
Wire-brush to clean, sound metal. Thin well-bonded paint is usually fine. Loose paint, rust scale, and mill scale are not — they trap air.
Gain creeping upward over months means the transmitter is turning up the volume to compensate for a weakening signal. That is couplant degradation, caught early. Put gain on the historian next to flow. It converts an outage into a maintenance ticket.
Not negotiable, and worth repeating because it is the most dangerous failure in the category.
The meter multiplies measured velocity by the full pipe area. On a half-full pipe it reports roughly double the real flow — confidently, with no error flag, and with a number that looks completely reasonable.
Avoid: high points that trap air, discharges that break the surface, gravity lines, and anything downstream of a control valve that can cavitate.
Prefer: vertical pipe with upward flow. Guaranteed full, no air pocket at the top, no sediment at the bottom.
If the pipe is not reliably full, stop. You need an area velocity meter, not a clamp-on meter, and no amount of installation skill will change that.
In order: (1) walk the line — a better location is often twenty feet away; (2) buy upstream length at the expense of downstream, since upstream disturbance matters far more; (3) switch to diagonal mode, which is less sensitive to profile asymmetry than reflect mode; (4) if you must install in a compromised location, use it for relative trending and document the condition. What you must not do is install in bad hydraulics and then treat the absolute number as truth.
Sometimes. If the mortar lining is well-bonded, yes. If it has delaminated — and on old pipe it very often has — there is an air gap inside the pipe that you cannot reach and cannot fill, and the signal stops there. There is no way to fix this from the outside. Test before you commit.
Yes, and it is often better than horizontal. A vertical pipe with upward flow is guaranteed full and has no air pocket at the top and no sediment at the bottom. Avoid vertical pipes with downward flow — they may not run full.
Look at signal strength, gain, and the quality index — not the flow rate. A flow reading with a weak or unstable signal is not a measurement. If gain is pegged at maximum, the instrument is straining, and you should re-couple, change mode, or move.
It depends almost entirely on pipe temperature. On an ambient-temperature line, years. On a hot line with the wrong grade of couplant, months. Trend the gain — a slow upward creep is the transmitter compensating for a weakening signal, and it is your early warning before the reading fails.
Yes, at the transducer locations. You need direct contact between transducer and pipe wall through couplant, and nothing else. Thin, well-bonded paint is usually acceptable; loose paint, rust scale, and mill scale are not.
Pipe material, wall thickness, lining, fluid, and a photo of the available straight run. We will tell you whether it will work — and if it will not, we will say so.
Request a quote