How to Verify a Guided-Wave Radar Level Reading
A guided-wave radar transmitter that disagrees with the tank gauge is one of the most common level complaints a technician gets handed. Before anyone reconfigures the transmitter, the honest first step is a careful manual gauge and a like-for-like comparison, because a large share of these discrepancies turn out to be unit and datum mismatches rather than instrument faults. This page walks through taking a manual reference reading, converting both numbers to the same basis, and deciding whether the radar actually needs attention.
Verify a GWR Level Reading in one line: To verify a guided-wave radar level reading, take a manual gauge of the tank with the level settled, convert the manual and transmitter readings to the same basis - both innage or both ullage, from the same reference point, in the same units - and compare them at the same moment. If they disagree beyond the tolerance the site allows, work through datum errors, configuration, and probe condition before suspecting the electronics. Opening a hatch on a pressurized or blanketed tank is a site-procedure decision, never a default step.
What You Need
The manual side needs whatever the site uses for reference gauging: a gauge tape with a plumb bob for innage or ullage, or a gauging stick on small atmospheric tanks. You also need the tank's reference records - the marked reference point on the hatch or gauge well, the reference gauge height, and the datum the strapping table is built from - plus a way to see the transmitter's live reading, either its local display or the SCADA tag, at the moment you gauge. If you are new to how the instrument itself measures, the explainer on guided-wave radar level is worth reading first.
The tank has to cooperate too. Gauge with the level settled, not during a fill or transfer, because the manual reading and the transmitter sample the surface at slightly different places and times and a moving level turns that into false disagreement. On blanketed, pressurized, or sour tanks, whether and how a hatch may be opened is governed entirely by site procedures and the permit system; many such tanks are gauged only through a gauge well or not manually at all, and that decision is not the technician's to improvise.
Take the Manual Reading Properly
Decide up front whether you are taking an innage, measured from the datum plate up to the liquid surface, or an ullage, measured from the reference point down to the surface, and record which one you took. Lower the bob slowly to avoid disturbing the surface, gauge from the same marked reference point every time, and take repeat drops until two consecutive readings agree closely; a single drop is an anecdote, not a reference. On tanks that carry a water layer under product, remember the radar and the tape may respond to different interfaces, so note any free water indication.
Read the transmitter at the same moment you read the tape, not after walking back to the control room. Level moves, and a comparison built from readings ten minutes apart on an active tank proves nothing. If a second person can call out the SCADA value while you gauge, or you photograph the local display timestamped, the comparison becomes defensible instead of approximate.
Compare on the Same Basis
Most "radar is wrong" reports die at this step. The transmitter reports level from its configured zero, which is set by its empty-distance or probe-length configuration, while the tape reports from the datum plate or the hatch reference point, and those origins rarely coincide. Convert both readings to the same basis - same origin, same direction, same units - before judging anything. An ullage compared against an innage, or a reading in feet compared against one in meters of product, produces exactly the kind of large stable offset that gets a healthy transmitter condemned.
Also compare level against level, not level against volume. If the operator's number came from a strapping table or an inventory system, it has been through an interpolation the transmitter knows nothing about, and a strapping-table error will masquerade as an instrument error. Pull the raw level the transmitter reports and the raw tape reading, and do the comparison there.
If They Disagree: Where the Error Usually Lives
A constant offset at all levels points at configuration or datum: the empty distance, the probe length, or the mounting-reference entry does not match the physical installation, or the two readings are simply on different origins. An error that grows or shrinks with level suggests the wrong span or a strapping mismatch. Erratic disagreement points at the process: buildup or coating on the probe, foam, an emulsion layer the radar tracks differently than the bob, or a dielectric change in the product weakening the reflection.
Only after those are ruled out does the instrument itself become the suspect, and by then the evidence tells you what to check: the echo curve in the configuration tool, the probe for damage or buildup at the flange, and the mounting nozzle. Resist the urge to trim the reading to match one tape drop; a trim that papers over a datum error guarantees the two references drift apart again at a different level.
Verifying the Result and Common Mistakes
The verification is complete when the transmitter and a properly taken manual gauge agree within the tolerance the site assigns at two or more well-separated levels, not just one. Agreement at a single level can be a coincidence of two errors crossing. Where the tank feeds custody transfer or inventory reconciliation, the comparison result belongs in the records; where the point is trended in a monitoring platform such as Merobix, the historical trend also shows whether the offset was stable or growing, which is exactly the context the next technician will want.
The recurring mistakes are all avoidable: comparing an ullage to an innage, gauging while the tank is moving, trusting one tape drop, using a kinked or repaired tape, comparing volume to level, and trimming the transmitter to force a match instead of fixing the reference. The manual gauge has its own uncertainty, so chasing a difference smaller than the tape can honestly resolve wastes everyone's time.
Frequently Asked Questions
How closely should a guided-wave radar match a manual tank gauge?
Within the tolerance the site assigns for that tank service, judged against the combined uncertainty of both methods rather than against zero. A manual gauge has real uncertainty from the tape, the reference point, and the operator, so perfect agreement is not the standard. What matters is that the difference is inside the site's acceptance band and stable across repeat checks; a small but growing offset is more significant than a slightly larger constant one.
Why does my radar read a constant offset from the tape at every level?
A constant offset at all levels is almost always a reference problem, not a sensor problem: the transmitter's empty-distance or probe-length configuration does not match the installation, or the manual reading and the transmitter are measured from different origins, such as an ullage compared against an innage. Fix the datum and the configuration entry rather than trimming the sensor, because a trim that hides a datum error will resurface as a level-dependent disagreement.
Can I verify a guided-wave radar without opening the tank?
Often yes. An independent reference can come from a sight glass or magnetic level gauge on a bridle, from a known transferred volume reconciled through the strapping table, or from watching whether the radar tracks a deliberate level movement accurately. Whether a hatch may be opened at all on a blanketed, pressurized, or sour tank is a site-procedure and permit decision, so plan the verification around the references the site actually allows.
Automation services
Need help turning this into a working system?
Merobix integrates SCADA, programs Allen-Bradley and Siemens PLCs, and designs and fabricates industrial control panels.
Meeting requests are reviewed before confirmation.