Loop checking: proving the instrument and the control system agree.

1. What a loop check actually proves

A loop check verifies a complete signal path: from the process connection at the field instrument, through the cable and junction box, into the marshalling and system cabinet, onto the I/O card, and up into the control system display, alarm and interlock logic. Every element in that chain has been installed and tested separately. The loop check is the first time they are proven together.

That is why loop checks are B-ITRs — dynamic records raised after energisation, feeding the commissioning phase rather than mechanical completion. Continuity and insulation resistance on the cable are static A-ITR checks and must be closed first.

2. Prerequisites before you start

Starting loop checks early is the classic false economy. The prerequisite set is short but absolute:

• Instrument calibrated with a valid certificate, and the calibration traceable to the tag — not the model.
• Cable tested — continuity, insulation resistance and screen integrity signed as A-ITRs.
• Terminations complete at field, junction box, marshalling and I/O, with the loop diagram as-built where it deviates.
• Control system configured — I/O address assigned, range and engineering units matching the datasheet, alarms configured.
• Power available and safe — cabinet energised, permit in place, and the loop's isolation status agreed with operations.

Half of all loop check failures are prerequisite failures rather than genuine defects. Enforcing the prerequisite list in the system that issues the record — instead of hoping the technician checks it — is the single highest-return change most projects can make here.

3. Calibration: what the record must contain

Calibration is a separate record from the loop check and must stand alone at audit. A defensible calibration record contains:

The test equipment certificate expiry is the detail that catches projects out. If a calibrator's certificate lapsed mid-campaign, every record it produced after that date is challengeable — which is why the calibrator register should be tracked with the same discipline as the instrument register.

4. The loop check sequence

1. Confirm the loop folder — loop diagram, instrument datasheet, calibration certificate, cable test records, and the blank loop test sheet.
2. Isolate correctly. Agree with the control room which interlocks and trips are inhibited, and record the inhibit so it cannot be forgotten at the end.
3. Inject at the field end. Simulate the process variable at the transmitter — pressure, current, resistance or frequency depending on type — at 0%, 25%, 50%, 75% and 100%.
4. Verify at every intermediate point where practical: junction box, marshalling terminal, I/O card LED or diagnostic.
5. Verify in the control system. Value, engineering units, scaling and sign are correct on the operator display and in the historian.
6. Test alarms and trips. Drive the signal through each alarm setpoint and confirm annunciation, priority and any interlock action.
7. Stroke the final element for control loops — valve travel, position feedback, fail position on loss of signal and loss of air.
8. Remove inhibits, restore setpoints, sign the record, and raise punch items for anything not corrected on the spot.

5. What actually fails, and what it means

Categorising these consistently matters more than it looks — see punch list categories. A loop with an unannunciated safety alarm is not a Category B item because the schedule would prefer it to be.

6. Managing loop checks at scale

A mid-size plant runs several thousand loops. At that volume the constraint stops being technical and becomes logistical: which loops are prerequisite-clear today, which cabinets are energised, which technicians are certified to sign, and which subsystem is closest to its gate.

• Sequence by subsystem, not by discipline. Loops are only useful when a whole subsystem's set is complete, because that is the unit that gets certified.
• Publish a daily prerequisite-clear list. Sending technicians to loops that cannot be checked is the largest single source of lost hours.
• Capture failures as punch immediately, linked to the tag and subsystem, so re-test scope is visible rather than remembered.
• Track re-test rate as a quality metric. A rising re-test rate normally means loop checks are being started before the prerequisites are genuinely closed.

This is exactly the sequencing that ITR management and the commissioning module are built to automate: the record knows its prerequisites, and the register knows which subsystem it is holding up.