Commissioning Loop Check Plan. What to Prepare Before the First Check.

Loop checking converts a wired plant into a commissioned plant. Every instrument, every cable pair, every PLC channel, every HMI tag gets exercised to confirm the path works end to end. Done well, the schedule holds. Done poorly, the schedule slips by weeks because every check turns into a debug session.

This covers the pre-check setup that makes the first day of loop checking productive, the sequence that scales, and the documentation discipline that keeps a large plant tractable.

Before the PLC is energized

Loop checking starts before you turn anything on. The pre-check list.

Tag database reconciliation. Every tag should have the same name in.

• The P&ID
• The I/O list
• The loop diagram
• The cable schedule
• The junction box schedule
• The PLC program
• The HMI, SCADA database
• The commissioning test sheet

In reality, naming drifts as each document is produced. Reconcile before the first loop check. A single correction memo or naming convention sheet distributed to every team member is cheaper than re-checking 50 loops later.

Calibration certificates. Every transmitter should have a current calibration certificate before it is wired. Commissioning teams that try to calibrate in the field at check time spend 3-4x the expected duration per loop.

Loop diagrams, actually drawn. Not "to be issued." Before your first commissioning day, every loop diagram you plan to check should be printed, reviewed, and on the technician's clipboard. Undrawn loop diagrams halt loop checks until someone in engineering can produce them.

Cable tagging matches documentation. Every installed cable should have a tag that matches the cable schedule. Commissioning teams that discover untagged cables during the check stop to trace, typically an hour per cable.

PLC program loaded and healthy. The controller should be in RUN, communicating, with all cards showing green LEDs. Loops cannot be checked through a non-communicating card.

Per-loop checklist template

The following checklist applies to a single loop from start to sign-off. Adapt the signal-specific sections to your signal class, see next section.

Spare I/O channel check. For any PLC channel configured but not wired to a field instrument, confirm the channel reads a stable value, typically 0 or open and is not alarming. Unverified spare channels create diagnostic noise during operations.

Per-signal-class procedures

Different signal types require different simulation equipment and accept different test stimuli.

4-20mA analog input. Use a loop calibrator, Fluke 705 or equivalent. Disconnect the field instrument at the marshalling terminal. Connect the calibrator in loop-powered or source mode and inject 4mA, 12mA, and 20mA. Verify PLC value at each point. Reconnect and verify the live instrument reading is in a plausible range for the process state.

RTD, PT100, PT1000 analog input. Use an RTD simulator or a precision decade resistance box. Disconnect the field sensor at the marshalling or I/O card terminals. Apply the resistance corresponding to 0°C, 100 ohms for PT100, a mid-range temperature, and the maximum range temperature. Verify PLC temperature value at each resistance. Reconnect. Note. RTD inputs are three-wire or four-wire. Confirm the wiring matches the card input type before applying simulation.

Thermocouple analog input. Use a thermocouple simulator, Fluke 714 or equivalent. Apply the millivolt signal corresponding to minimum, mid-range, and maximum temperature for the thermocouple type in use, J, K, T, etc. Verify PLC value at each. Cold junction compensation in the I/O card must be active for the reading to be correct.

HART transmitter. Use a HART communicator, Emerson 475 or equivalent. For the signal portion, a standard 4-20mA injection confirms the analog path. For HART device data, device status, diagnostics, secondary variables, poll the device over HART from the communicator or from a HART multiplexer. Verify the primary variable, device tag, and configured range in the HART device match the I/O list. For HART-enabled cards with a host system reading device variables, verify the HART variables appear in the PLC program as expected.

Foundation Fieldbus, FF. FF devices are addressed and configured digitally, not by analog signal. The loop check for FF consists of. Confirm device is detected on the segment. Verify device address matches the segment design. Confirm function block configuration, AI, AO, PID matches the control narrative. Verify the live process value appears in the host system. Verify alarms are mapped. There is no simulated signal injection. The device must be live on the process or have a simulation mode enabled in the device configuration tool.

Digital input, dry contact or proximity. Open and close the contact by hand or by actuating the field switch. Verify the PLC input changes state and the HMI display reflects the correct status, open, closed, run, stop, etc. For a process switch, pressure, level, temperature, bypass or isolate, then simulate the contact closure with a jumper at the marshalling terminal.

Digital output, discrete relay or solid-state. Force the PLC output from the commissioning interface. Verify the output energizes and the field device, solenoid, motor starter, valve actuator responds. Confirm position feedback, if present reads correctly. Remove the force. Verify the output returns to its normal de-energized state.

Analog output, 4-20mA to control valve or VFD. Force the PLC output to 0%, 50%, and 100% from the commissioning interface. Measure the milliamp signal at the output terminals with a calibrated meter. Observe valve position or VFD speed at each command. Confirm position feedback reads back correctly at each step.

SIS loop proof tests

Safety Instrumented Function proof tests are distinct from basic loop checks and require separate sign-off. The procedure per IEC 61511 clause 16 includes.

1. Confirm the SIF is documented in the Safety Requirements Specification with SIL target, voting architecture, and proof-test interval.
2. Notify operations that the safety function will be taken out of service for proof testing. Follow the management of change procedure for temporary bypass.
3. Simulate the initiating condition at the field instrument, pressure, temperature, or level transmitter by applying the trip setpoint value using a calibrator or process simulation.
4. Verify the final element, shutdown valve, motor trip, deluge valve responds within the required response time.
5. For voted architectures, 1oo2, 2oo3, test each sensor independently and verify the voting logic in the SIS controller functions correctly for each combination of sensor states.
6. Restore the final element to the normal position and confirm the process is safe to receive the instrument back.
7. Sign off with both the commissioning engineer and the operations representative. The proof-test record becomes part of the SIS maintenance history and must be retained for the life of the plant.

Do not combine the SIS proof test with the basic loop check on a single form. Regulators and auditors expect separate documentation for safety loop verification.

The loop check sequence

Check one loop end to end before starting the next. The sequence for a single analog input loop.

1. Isolate the loop at the marshalling strip or the transmitter terminal
2. Apply a simulated signal, loop calibrator at 4mA and 20mA, or range endpoints
3. Confirm the correct value appears in the PLC raw register, 0%, 100%
4. Confirm the scaled value appears in the engineering-unit tag, e.g. 0-100 GPM
5. Confirm the HMI displays the scaled value
6. Confirm the historian trends the value
7. Exercise the alarm limits by stepping input above, below alarm thresholds
8. Confirm the alarm annunciates on the HMI and any alarm historian
9. Remove the simulation and restore field wiring
10. Sign the loop-check form

For a digital input or output loop, the sequence is similar but uses contact closure or coil energization instead of analog simulation.

The paperwork that survives an audit

Each loop-check form should capture.

• Tag number
• Date and time of check
• Technician name and signature
• Engineer witness name and signature
• Test value applied at each step
• Indicated value on the HMI
• Pass, fail per step
• Any deviations or temporary workarounds
• Final sign-off

Deviation log. Every failed step or non-conformance gets a deviation entry. Tag number, step that failed, observed value, expected value, corrective action taken, and resolution sign-off. The deviation log is the audit trail that shows every problem was found and fixed before handover.

Hand-off package. At the end of loop checking, the commissioning package handed to operations should include. Completed loop-check forms for every tag, deviation log with all items closed, as-built I/O list with PLC addresses confirmed, calibration certificates, and proof-test records for SIS loops. The package should be organized by loop number or tag number to allow random sampling during the handover audit.

Common delay patterns

Tag database not reconciled. Every loop takes 2-3x the expected duration because every check includes a "wait, what is this actually called" step.

Missing loop diagrams. Technicians stop while engineering produces the diagram on-the-fly.

Calibration shortcuts. Transmitters that were not bench-calibrated need to be calibrated at the marshalling strip, doubling the loop-check duration.

Signal class mismatches. An analog input wired to a digital card passes zero loop checks until rewired. These should be caught during instrument installation, but they are routinely caught at loop check.

SIS tags mixed with BPCS tags. The SIS proof-test procedure is different from the basic loop check. Plants that try to use a single procedure for both fail the SIS verification audit.

Where the I/O list factors in

Every loop-check form needs a correct tag, signal class, HMI reference, and connected equipment. Those fields come from the I/O list.

If the I/O list was built by hand from a P&ID pile, expect a 5-10% error rate that surfaces during loop checking. Misspelled tags, wrong signal class on flagged cases, FCV classified as DO instead of AO, missing rows. Every error found during commissioning is a project delay.

Whatever tool produced the I/O list, check it against this column set before handing the spreadsheet to commissioning. Tag Number, Signal Class, Description, Connected Equipment, P&ID Reference, Page Number. Missing any of those means the loop-check form cannot be filled in correctly. The I/O list creation guide covers how to build and verify that column set before the list reaches commissioning.

A structured I/O list derived from P&ID extraction also provides the loop count per signal class needed for scheduling. If you know you have 180 analog inputs, 42 analog outputs, 310 digital inputs, and 88 digital outputs, you can assign technician time realistically before the first day of checks rather than discovering the scope mid-way through.

Related

• P&ID to I/O list
• PLC tag naming conventions. Reconcile before loop check
• Loop diagrams vs P&IDs. Which document is authoritative for what
• Estimating PLC I/O count
• 4-20mA vs HART vs fieldbus. Signal path detail relevant during loop check
• Commissioning lifecycle
• Operations handover

FAQ

How long does a typical instrument loop check take.

Once the pre-check paperwork is in order, a simple analog loop takes 15-30 minutes. Complex loops, redundant transmitters, fieldbus devices, SIS loops with proof-test requirements take 1-4 hours. A 500-point plant typically takes 6-10 weeks of loop checking with a two-technician team.

What documents do I need before starting loop checks.

I/O list with current PLC addresses, loop diagrams, ISA 5.4, cable schedule, junction box schedule, calibration certificates for transmitters, and the commissioning test procedure. Missing any one of these stops the loop check cold.

What is the biggest cause of loop-check delays.

Mismatched tag databases. The field instrument says FIT-101, the HMI says FIT_101, the PLC program says FT_101, and the commissioning spreadsheet says Flow_01. Each naming variation looks fine in isolation. Together they fail every auto-match. Fix the naming convention before energizing.

Should I loop-check before or after setting the alarms.

Before. Alarm limits are commonly set during loop check by stepping the input across range and watching the HMI behavior. Setting alarms upstream of the loop check leads to nuisance trips during the check itself.

How do I sequence loop checks across a plant.

By control loop, not by signal class. All analog inputs first, then all discretes might seem efficient, but it leaves every loop half-checked. Check a loop end-to-end, transmitter to HMI to control action one at a time. It is slower in unit terms, faster in project terms.

What do I do when a spare I/O channel is not associated with any field device.

Confirm the channel is in a known safe state, typically reading 0 or open, not alarming. Tag the channel in the PLC program as "spare" with the physical card and channel reference. Document it on the I/O list so future instrument additions can reference available spares. Untagged spare channels create support problems when the next instrument is added.