Level Measurement Instrument Calibration: A Complete Guide

Most plants run more than one type of level instrument. The same site might have ultrasonic sensors on open tanks, radar transmitters on pressurized vessels, and magnetic level gauges on interface applications - and each one gets calibrated a little differently. That's fine on paper, but in practice it's often the reason calibration gets inconsistent: technicians default to whatever procedure they remember best, and the instrument that doesn't get much attention is usually the one that drifts furthest before anyone notices.

This guide pulls the three most common level measurement technologies - ultrasonic, radar, and magnetic - into one reference. It covers what each one has in common, where the calibration steps diverge, and how to build a maintenance routine that doesn't depend on one technician's memory. If you need the full step-by-step procedure for a specific instrument type, we've linked out to the detailed guides at each section.

Why a Calibration Strategy Matters, Not Just a Calibration Procedure

Individually, calibrating one sensor is straightforward. The harder problem shows up at the plant level: with a dozen or more instruments across three different technologies, inconsistent calibration intervals and undocumented procedures are what actually cause inventory discrepancies, false alarms, and unplanned inspections. A strategy - standard intervals by criticality, a shared documentation format, and a common understanding of what "good" looks like for each technology - matters more than any single calibration being done perfectly.

Level Measurement Technologies at a Glance

Before getting into calibration specifics, it helps to know what's actually being measured and why the procedure differs.

Technology	Measurement Principle	Contact With Media	Typical Calibration Basis
Ultrasonic	Time-of-flight of an acoustic pulse through air	Non-contact	Empty/full distance, echo processing profile
Radar (non-contact)	Time-of-flight of a microwave pulse	Non-contact	Empty/full distance, dielectric constant of medium
Guided wave radar	Microwave pulse traveling along a probe	Contact (probe only)	Empty/full distance along probe length
Magnetic level gauge	Float position read through a sealed chamber	Contact (float only, media-isolated)	Mechanical scale alignment, float specific gravity

The common thread: every one of these technologies needs a defined zero reference and a defined full-scale reference. The difference is in what generates the signal and what can interfere with it.

Ultrasonic Level Gauge Calibration

Ultrasonic calibration centers on setting the zero (empty) and span (full) distances, then verifying against an independent reference at a few intermediate points. The main variable to manage is the acoustic path - anything that disturbs the surface (foam, turbulence) or blocks the beam (internal obstructions) will throw off readings even when zero and span are set correctly.

Read the full procedure: How to Calibrate an Ultrasonic Level Gauge - covers dry vs. wet calibration, dead-band considerations, environmental compensation, and a field example of a foam-related drift issue that looked like a calibration problem but wasn't.

Ultrasonic Level Transmitter

Radar Level Sensor Calibration

Radar calibration follows the same zero/span logic as ultrasonic, but with one added variable: the dielectric constant of the process medium. Because radar measures via electromagnetic reflection rather than sound, weakly reflective media - light hydrocarbons, liquefied gases, low-conductivity liquids - return a weaker echo, and the transmitter needs to know the medium's dielectric value to interpret that echo correctly.

Typical calibration steps:

Confirm empty and full reference distances from the vessel drawing, same as ultrasonic - measured from the antenna face to the lowest and highest points in the measuring range.
Enter the dielectric constant for the process medium, if the transmitter supports it. Most manufacturers publish reference tables; guessing this value is one of the most common causes of poor accuracy on low-dielectric media.
Dry calibration uses a reference target (a flat metallic plate, typically at least a meter square) positioned at the known empty and full distances, with the antenna kept perpendicular to it.
Wet calibration fills the vessel to known levels and adjusts the output to match - necessary when internal obstructions generate false echoes that need to be mapped and filtered out.
False-echo mapping, sometimes called an "echo curve" or "empty curve," records the reflections from fixed internal structures (agitators, nozzles, ladders) so the transmitter's software can ignore them during normal operation.
Verify at intermediate levels, not just the endpoints - particularly important on vessels with a non-flat roof or angled internals, which can distort the reading near full.

Radar tends to require less recalibration over time than ultrasonic, since it isn't affected by temperature-driven changes in the speed of sound. It's a common upgrade path for applications where ultrasonic struggles - heavy vapor, foam, or dust - since electromagnetic waves aren't attenuated by those conditions the same way sound waves are.

80GHz Radar Level Sensor

Magnetic Level Gauge Calibration

Magnetic level gauges work differently from the other two. The visual indicator - a column of flipped magnetic flags - tracks the physical position of a float and generally doesn't require calibration in the same sense; there's no signal processing to adjust. What does require attention is the scale alignment and, where a transmitter is attached to the gauge, calibration of that transmitter's output.

Typical checks and procedure:

Verify the float's specific gravity matches the process medium. A float sized for the wrong density will ride at the wrong position even if everything else is installed correctly - this is a specification issue, not something you can fix by adjusting the scale.
Check free movement of the float in the guide chamber. Any sticking, caused by debris, corrosion, or a bent chamber, will show up as a level reading that lags behind or doesn't move at all.
Align the scale using the manufacturer's correction procedure, which typically accounts for the offset between the float's magnet centerline and the base of the float.
If a magnetostrictive or reed-chain transmitter is attached, perform a standard zero/span calibration on that transmitter: move the float to the 0% reference, confirm or trim the output, then repeat at 100%.
Cross-check the visual column against the transmitter output periodically - a mismatch between the two usually points to a transmitter calibration drift rather than a mechanical float problem, since the visual indication itself has no electronics to drift.
Inspect for magnetic interference if the readings look consistently off - strong external magnetic fields near the chamber can affect flag response, though this is uncommon in typical installations.

Because the core visual indication is mechanical rather than electronic, magnetic level gauges tend to hold their accuracy over long periods with minimal intervention - the main maintenance burden is on the float chamber and, where fitted, the transmitter.

Side Mounted Magnetic Level Gauge

Dry Calibration vs. Wet Calibration: The Same Choice, Every Technology

Regardless of which instrument you're working with, the same fundamental decision applies: calibrate using known reference distances (dry) or calibrate against an actual filled vessel (wet).

Dry calibration is faster and doesn't require taking a vessel offline or filling it to specific levels. It's the default for general process monitoring where a small margin of error is acceptable.
Wet calibration accounts for the real behavior inside that specific vessel - internal obstructions, non-standard geometry, actual medium properties - and is the right choice for custody transfer, regulatory reporting, or any application where the cost of being wrong outweighs the cost of the extra setup time.

A reasonable middle ground, used often in practice: dry-calibrate at commissioning using accurate drawings, then verify with a single wet check against a dip tape or sight glass reading before the instrument goes into full service.

A Field Example: Same Symptom, Three Different Root Causes

A processing facility running all three instrument types across different vessels once traced a recurring "the level reading doesn't match the tank" complaint back to three unrelated causes on three different sensors, all reported the same week.

On the ultrasonic sensor, the cause was a foam layer confusing the echo - already covered in the ultrasonic guide linked above. On the radar unit, the cause was a missing dielectric constant setting; it had been left at a default value appropriate for water, but the actual medium was a much less reflective solvent, weakening the echo and introducing noise near the low end of the range. On the magnetic level gauge, the float itself was the problem - a replacement float installed during a previous repair had the wrong specific gravity rating for the process medium, so it was riding lower than it should have, regardless of how well the scale was aligned.

None of the three were fixed by "recalibrating" in the generic sense. Each required understanding what that specific technology depends on to produce an accurate reading in the first place. That's the practical argument for treating calibration as instrument-specific knowledge rather than a single universal checklist.

How Often Should Each Instrument Type Be Recalibrated?

Instrument	Custody Transfer / Regulatory	Standard Process Monitoring	Harsh or Variable Environments
Ultrasonic	Quarterly verification	Annual check	More frequent spot checks recommended
Radar	Quarterly verification	Annual check	Generally more stable; verify after any medium change
Magnetic level gauge	Annual scale/float inspection	Inspect at turnaround intervals	Check float and chamber condition more often in corrosive service

Manufacturer-specified intervals should always override general guidance where they're more conservative, particularly for any instrument feeding a safety-instrumented system.

Common Mistakes That Cut Across All Three Technologies

Treating every discrepancy as a calibration problem. As the field example shows, the fix is often a configuration setting, a mechanical component, or a process condition - not the zero/span values themselves.

Skipping documentation. Whichever technology is involved, a calibration record that only says "calibrated, OK" is close to useless for spotting drift trends. Record the reference method, ambient conditions, and before/after values every time.

Assuming factory settings are final. Every instrument ships with a default configuration that may not match your specific vessel geometry or process medium. Verifying that configuration during commissioning - not just after a problem shows up - prevents most of the issues covered in this guide.

Inconsistent intervals across similar risk levels. If a custody-transfer application on one technology is checked quarterly, an equally critical application on a different technology at the same site should be on a comparable schedule, not left to whatever the original commissioning contractor happened to set.

Building a Simple Calibration Program

For a site running mixed instrument types, a workable program usually has three parts:

A criticality tier for every level instrument - custody transfer/safety, standard process, and non-critical - with a defined interval for each tier, applied consistently regardless of technology.
A shared calibration record format, so that whoever pulls the maintenance history later can compare instruments side by side instead of interpreting three different logging conventions.
A short technology-specific checklist for each instrument type - pulled from the sections above - so technicians aren't relying on memory or applying an ultrasonic mindset to a magnetic gauge.

None of this requires special software or a large budget. A shared spreadsheet with the three checklists attached covers most small-to-mid-sized operations adequately; larger sites with dozens of instruments typically fold this into their existing CMMS.

Final Thoughts

Ultrasonic, radar, and magnetic level gauges solve the same basic problem in three different ways, and that difference matters more at calibration time than most maintenance schedules account for. Standardizing intervals and documentation across a site is useful, but it shouldn't paper over the fact that each technology has its own failure modes - foam on ultrasonic, dielectric mismatches on radar, float specification errors on magnetic gauges. Knowing which one you're dealing with, and what that specific instrument actually depends on for accuracy, is what separates a quick fix from a repeat service call.

References

Instrument Society of America (ISA) - general guidance on level measurement instrumentation and calibration practices.
American Petroleum Institute, Manual of Petroleum Measurement Standards, Chapter 3 - tank gauging and level measurement principles.
National Institute of Standards and Technology (NIST) - reference data on the speed of sound in air and its temperature dependence.
Manufacturer technical documentation for ultrasonic, radar, and magnetic level instruments - dead-band, dielectric constant, and float specification references.
Not sure which level measurement technology fits your application, or need help auditing calibration practices across a mixed instrument fleet? Send us your vessel details and process conditions - we're happy to help you work through it.