| Reading is stable but consistently wrong | Wrong meter factor entered, nominal pipe size used instead of the real inner diameter, or the wrong density and reference conditions | Check the factory data plate against the transmitter settings and enter the measured inner diameter, not the nominal one |
| Signal jumps and the loop hunts | Partly filled pipe, entrained air, a pulsating positive displacement pump, or a meter sized far too large | Move the meter into a rising line or a low point, fit an air release upstream, add a pulsation damper, and check the meter is not running below a tenth of its range |
| Electromagnetic meter reads zero or wanders | Conductivity too low, empty or partly filled pipe, coating on the electrodes, or poor earthing | Verify the conductivity, guarantee a full pipe, clean the electrodes or switch to a capacitive type, and fit earthing rings or a proper bonding strap |
| Coriolis zero drifts after commissioning | Zeroed while flow was still passing, stress from the pipe supports, or a large temperature change since zeroing | Close both isolation valves, let the temperature settle, re-zero, and check that the meter carries no pipe weight or misalignment |
| Clamp-on ultrasonic loses signal | Air gap or dried couplant, a rough or corroded pipe wall, a liner thicker than entered, or gas bubbles in the liquid | Re-apply couplant, move to a clean section of pipe, enter the real wall and liner data, and stay away from the downstream side of a pump or valve |
| Turbine meter accuracy fades over months | Bearing wear, deposits on the rotor, or debris strikes in an unprotected line | Fit a strainer upstream, set a recalibration interval, and consider a meter without moving parts if the liquid stays dirty |
| Batches come out short or long | Valve closing time not compensated, damping set too high, or the flow ramped down too late | Set an in-flight compensation for the valve, lower the damping, and finish the batch with a slow dribble stage |