YARDPUMP · Technical Article

Why Does a Pump Test Magmeter Indicate Flow with the Water Stopped?

Establish a full-pipe zero condition, then separate real residual flow, entrained air and electrical interference before adjusting a test meter.

In a pump acceptance rig, the motor has stopped and the display still flickers a few litres per minute. That is a reason to check the test setup before moving a pass or fail boundary. A stopped motor does not prove zero flow through the meter. Start by demonstrating a known full, stationary condition in the measuring section. Only then investigate zero offset, unstable electrodes or electrical noise.

A motor at rest does not settle the hydraulic question

Where the piping has an elevated line, a return loop or a connected vessel at a different level, water may continue moving slowly after a pump stops. A check valve can leak. A meter near a section that partially drains may have an air pocket even if a nearby pipe appears full. Some magmeter models report empty pipe, but coverage depends on configuration and the electrode condition. Small readings that decay toward zero can reflect a real hydraulic event as well as damping in the transmitter.

An electromagnetic flowmeter measures a signal generated as electrically conductive liquid moves through a magnetic field. It needs sufficient conductivity, liquid contact at the electrodes and the manufacturer's installation, grounding and signal wiring. A noisy reference or an exposed electrode can produce an unstable reading even at rest. Nearby power equipment, cabling or a drive may contribute to the displayed behaviour, but noise should be demonstrated by a controlled change rather than inferred from proximity alone.

Establish an independent zero and compare signals

A qualified technician should follow the meter and rig instructions for a full pipe with no velocity. Use the approved isolation of a flooded section, without running the pump against closed valves. Check whether pipes are safely supported and whether test valves or other sources can pass flow into the section. Match water temperature and conductivity to the test duty where relevant. If safe and available, compare a separate low-flow reference or tank balance. Never crack open a pressurized connection or defeat protective earthing to check a theory.

Log the status indicators and compare the raw velocity or flow display, any filtered signal and the totalizer response if the instrument makes these available. A damping setting can prolong a displayed reading after motion stops. Low-flow cutoff can make a readout look quiet while the underlying measurement is noisy. Save the original settings and avoid changing them mid-test. After proving hydraulic zero, compare conditions that leave the measuring section stationary while surrounding equipment changes state under a controlled approved procedure. Use suitably qualified personnel for grounding, shielding or wiring checks and isolate power before altering equipment.

For an illustrative example, assume a sealed independent full-water section shows a stable 0.2 L/min offset with all sources steady, and the same section flickers between positive and negative values only when adjacent drives operate. The repeatable effect is evidence for investigating the signal installation. The values are hypothetical and are not permissible tolerances. If the section cannot be proved stationary, the same display pattern does not settle whether some water was moving.

Return to acceptance only after the meter is fit for the duty

A stable offset under a valid documented zero may be treated by the approved instrument zero verification or adjustment procedure. A drifting or state-dependent reading calls for inspection of fill, electrode condition, grounding and signal paths before adjustment. Keep evidence of the old settings and retest a qualified low-flow point as well as the normal duty. A clean zero alone cannot prove accuracy at nonzero flow.

For a purchasing test, state how the zero condition was made, how long it was observed, and which raw and reported channels were compared. When the uncertainty is material to the acceptance threshold, do not use the reading to reject a pump until the measuring setup has been confirmed.

Frequently asked questions

Can an empty pipe be a convenient zero check?

No. A dry electrode is a different condition. Use the documented full and still test method.

What if the number disappears below the low-flow cutoff?

That can conceal small movement or noise. Check the underlying indication and approved settings.

Does reversing the flow direction change what counts as zero?

True zero is stationary liquid. The sign convention matters for small nonzero indications and for checking whether a backflow path exists.