80 GHz Radar Level Commissioning: From Geometry to Acceptance

Commissioning an 80 GHz radar is not finished when a plausible level appears on the display. The work must prove that the installed geometry matches the configuration, the intended surface is tracked throughout the useful range, the signal reaches every receiving system correctly and abnormal states produce the agreed response. A narrow beam can simplify difficult vessels, but it cannot correct an undocumented reference plane, a recessed antenna or incorrect PLC scaling. Acceptance should leave a repeatable baseline for operations and maintenance.

Complete mechanical inspection first

Match the model, antenna, process connection and accessories to approved documents. Measure nozzle bore and length, antenna projection, alignment, empty distance, maximum-level clearance and surrounding structures. Photograph the final assembly. Close defects affecting the beam, seal, pressure boundary, hazardous compliance or access before changing signal parameters; otherwise configuration may conceal an installation problem.

Use one reference datum

Define the antenna reference plane and map it to tank zero, empty and full levels. Check whether the control system expects distance, level, percent or volume. Include adapters and extensions in dimensions. Record the independent reference and its uncertainty. A common datum prevents offsets from being added separately in the radar, PLC and tank table and makes later comparisons meaningful.

Review the raw echo before suppression

Capture the echo curve with the tank at a known condition and identify the product surface, nozzle and fixed structures. Check signal-to-noise margin and quality, not only the selected distance. Investigate unexpected strong echoes physically. If false-echo mapping is necessary, document the level and process state used and prove that a real high-level return will not be suppressed.

Set range and dynamics from the process

Configure empty and full distances, blocking distance, response rate, damping and lost-echo behavior from actual filling and emptying rates. Avoid excessive damping that delays alarms or masks instability. Confirm behavior during agitation, inlet flow, foam, dust or condensation where relevant. Save both default and as-left parameters so every departure has an identified purpose.

Verify every output path

For 4–20 mA, inject or establish several known values and compare transmitter current, input counts, engineering value, HMI and historian. For HART, Modbus or fieldbus, verify variable assignment, units, data type, byte order, quality and timestamp. Check that only one system applies linearization and that the displayed direction matches the physical level.

Test alarms and failure states

Test high and low alarms near their setpoints, including hysteresis and delay. Verify open circuit, power loss, lost echo, device fault and communication interruption. Confirm the PLC or DCS distinguishes invalid or stale data from a real low level. Remove every force and bypass under independent check and prove final equipment action where the approved procedure permits.

Observe representative operation

Run the measurement through a meaningful fill, empty or batch cycle and compare it with an independent reference at useful points. Correlate any instability with process state and raw diagnostics. If commissioning cannot include foam, maximum rate or near-full conditions, record those gaps and schedule an early-service review rather than declaring the entire operating envelope proven.

Build the as-left package

Retain model and serial number, firmware, dimensions, photographs, configuration, echo curves, output test, alarm results, open items and approvers. Store files under the equipment tag where another technician can retrieve them. The package should define the approved application envelope and trigger review after product, geometry, range, firmware or control-logic changes.

Engineering checklist

Inspect mechanics before parameter changes.
Use one documented reference datum.
Capture and interpret the raw echo curve.
Set dynamics from real process rates.
Test analog or digital paths end to end.
Retain a complete as-left baseline.

Frequently asked questions

Is auto setup enough for commissioning?

No. Automatic setup does not verify geometry, system scaling, alarms or fault behavior.

When should false-echo mapping be used?

Only after physical causes are reviewed and with evidence that valid process echoes remain detectable.

Why keep the first echo curve?

It provides a known-good comparison for buildup, mechanical change and later fault diagnosis.

Need a project-specific review? Send process data, drawings, photographs and acceptance criteria through our contact page.