What Is Loop Calibration?
Loop calibration is how an instrument technician proves that a measurement is trustworthy from the sensor all the way to the number an operator sees. Rather than checking a transmitter alone, it verifies the whole signal path - element, transmitter, wiring, and the controller input - as one system. This guide explains what loop calibration is, how the As-Found/As-Left method works, and why oil and gas facilities depend on it.
Loop Calibration in one line: Loop calibration is the end-to-end verification of a complete instrument loop - applying a known input at the sensor and confirming the value that appears at the controller or SCADA display matches within tolerance, correcting the transmitter or scaling if it does not.
Instrument Calibration vs Loop Calibration
A bench or single-instrument calibration checks one device in isolation: apply a known pressure or temperature to a transmitter and confirm its output current is correct. Loop calibration is broader - it exercises the entire chain from the primary element through the field wiring, junction boxes, and analog input card, up to the value the control system reports. A transmitter can be perfectly calibrated on its own yet still show a wrong reading on the screen because of a scaling error in the controller, a lead-wire fault, or a mismatched input range.
By stimulating the loop at the process end and reading the result at the display end, the technician catches errors that a device-level check would miss. This is why commissioning and periodic proof-testing in oil and gas rely on documented loop calibrations rather than isolated instrument checks.
The As-Found / As-Left Method
Calibration is documented in two passes. The As-Found reading captures the loop's performance before any adjustment - it reveals how far the instrument has drifted since the last check and feeds reliability and audit records. The technician then makes corrections and records the As-Left values, which prove the loop now reads within its allowed tolerance.
Typically five points are tested across the range - 0, 25, 50, 75, and 100 percent - checked in both rising and falling directions to expose hysteresis and non-linearity. Each point compares the applied input against the reported value, and the error must fall inside the tolerance band, often a fraction of a percent of span for custody or safety-critical loops.
Where Loop Calibration Fits in Oil and Gas
Regulated measurements - custody transfer meters, environmental monitoring, and safety-instrumented functions - carry mandated calibration intervals, and the loop calibration record is the evidence that the reading is defensible. A drifted pressure or level loop can misreport inventory, trip a shutdown falsely, or fail to trip when it should.
Loop calibration happens at the field and controller layer, below any SCADA platform. A cloud SCADA such as Merobix reads the already-calibrated value the controller produces; it reports and trends the number but does not itself perform or replace the physical calibration. Accurate loops upstream are what make the dashboard trustworthy.
The Equipment on Each End of the Test
Most loop calibrations are run with a documenting process calibrator - a single instrument that sources the process input and measures the resulting current. For a pressure loop that means a hand pump or pressure controller teed into the transmitter's test connection; for temperature loops the sensor is usually disconnected and an RTD or thermocouple simulation injected in its place. Because simulation bypasses the element itself, the sensor still deserves its own periodic check, which is what the guide on how to bench test an RTD covers in detail.
The other end of the test is a person at a display. Someone must confirm what the controller and HMI actually show at each test point, because the whole purpose of a loop calibration is to prove the displayed value, not just the transmitter output. On a smart transmitter, a HART communicator distinguishes between a sensor trim, which corrects the measurement itself, and a re-range, which only changes how the measurement maps onto the 4-20 mA current loop. The calibration record should state which adjustment was made, because the two mean very different things for traceability.
A Worked Five-Point Example
Take a transmitter ranged 0 to 100 in whatever engineering unit the loop uses. The 4-20 mA standard fixes the expected current at each test point regardless of the unit: the signal spans 16 mA, so each percent of span is worth 0.16 mA.
| Percent of span | Applied input | Expected current |
|---|---|---|
| 0 | 0 units | 4 mA |
| 25 | 25 units | 8 mA |
| 50 | 50 units | 12 mA |
| 75 | 75 units | 16 mA |
| 100 | 100 units | 20 mA |
At each point the technician records the applied input, the measured loop current, and the value the control system displays. Error is normally expressed as a percent of span: subtract the expected current from the measured current and divide by the 16 mA span. If the applied input is 50 units and the loop reads 12.2 mA, the error is 0.2 divided by 16, or 1.25 percent of span - acceptable or not depending on what the calibration data sheet allows for that loop. The tolerance itself comes from the site's calibration program, not from the instrument.
Subtle Faults the Five Points Expose
Beyond outright scaling and wiring errors, the pattern of the five points tells its own story. A constant offset at every point suggests a zero shift or a ground loop injecting a fixed error; an error that grows toward full scale points to a span problem; a rising pass that disagrees with the falling pass reveals hysteresis or a heavily damped transmitter that was not given enough settle time between points. Moisture in a junction box shows up as an error that changes with the weather, and a transmitter landed on the wrong analog input channel produces a live, plausible display value that belongs to a different instrument entirely.
Full-scale behavior deserves particular attention. At 20 mA the loop drops the most voltage across its wiring and input resistance, so a marginal power supply or an overloaded loop can read correctly at low current yet clip near the top of the range. That failure hides from any single-point check and from most bench tests - it only appears when the loop is driven to 100 percent through its real wiring, which is exactly what the five-point loop calibration does.
Preparing a Live Loop Safely
Most calibration mistakes happen before the calibrator is connected. A loop that feeds an interlock, an alarm, or a control valve will act on test signals unless it is properly bypassed, so preparation follows the site's permit system, with qualified personnel making the isolation decisions.
- Confirm the loop's function and what it drives - control, alarm, shutdown, or reporting only.
- Obtain authorization and apply bypasses or forces per the site's procedures.
- Notify the control room so operators expect the moving values.
- Isolate the process connection per procedure before opening any impulse line.
- After the As-Left pass, remove every force and bypass, verify the live reading agrees with the process, and close out the record.
Closing the record matters as much as the numbers in it. As-Found drift history is the raw material for calibration interval optimization - loops that never drift can earn longer intervals, while repeat offenders justify shorter ones or a hardware fix.
Frequently Asked Questions
What is the difference between loop calibration and instrument calibration?
Instrument calibration checks a single device on its own. Loop calibration verifies the whole signal path - sensor, wiring, transmitter, and controller input - as one system, catching scaling and wiring errors that a device-only check cannot see.
What does As-Found and As-Left mean?
As-Found is the loop's error recorded before any adjustment, showing how much it drifted. As-Left is the error recorded after correction, proving the loop now reads within tolerance. Both are documented for audit and reliability records.
How often should a loop be calibrated?
It depends on the loop's criticality and regulations. Custody-transfer and safety-instrumented loops follow mandated intervals; general process loops are often calibrated annually or based on observed drift. Facilities set intervals in their calibration management program.
Does a smart HART transmitter still need a loop calibration?
Yes. The digital electronics can be healthy while the analog output, wiring, power supply, or controller scaling is wrong, and none of those are proven by a device-level check. A loop calibration validates the full path to the displayed value. HART tools help by separating sensor trim from re-ranging, but they complement the loop check rather than replace it.
What happens when the As-Found values are out of tolerance?
The technician records the failure, adjusts or repairs the loop, and documents passing As-Left values. The out-of-tolerance finding is then reviewed under the site's calibration program: how long the loop may have been reading wrong, whether affected measurements need assessment, and whether the calibration interval should shorten. For custody or safety loops that review follows the site's defined procedures.
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.