A technically valid thickness reading is not automatically ready for a Fitness-for-Service assessment. The engineer must also know what was measured, where it was measured, what area the value represents, how the reading was obtained and how uncertainty was controlled.
This distinction matters when inspection identifies corrosion or local metal loss. A spreadsheet may contain hundreds of values and still leave the damage geometry unclear. Conversely, a smaller controlled dataset may be useful when its locations, coverage, method and limitations are traceable to the engineering question.
This guide explains how to judge the quality of thickness and mapping data before it enters an FFS workflow. It does not prescribe a universal grid, a fixed number of readings or an acceptance limit. Those decisions depend on the applicable procedure, equipment geometry, expected damage morphology, assessment route and project specification.
Key Takeaways
• Measurement accuracy is only one part of data quality. Identity, location, coverage, context and traceability also affect whether the data can support an engineering decision.
• A minimum thickness value may identify a concern, but it may not define the length, width or profile of local metal loss.
• There is no universal thickness-mapping grid. Coverage must be selected around the damage question and the assessment needed.
• Outliers and no-read locations should be repeated and investigated, not silently removed from the dataset.
• The practical outcome is one of three routes: proceed to engineering review, use the data with explicit limitations, or perform targeted reinspection.
A Thickness Reading Is Not Automatically Decision-Ready
A reading becomes useful engineering evidence when it can be connected to the component, the exact location, the area represented, the measurement process and the uncertainty that remains. A number without that chain may be correct at the probe position but insufficient for the damage model.
For example, a value of 7.8 mm says little by itself. The FFS engineer may need to know whether it came from the shell, head, nozzle neck or weld-adjacent area; whether it represents a single spot or the minimum within a mapped region; and whether the measurement was taken through coating, on a rough surface or at elevated temperature.
A practical review can use six dimensions:
| DimensionQuestion to answer | |
|---|---|
| Equipment identity | Can the value be tied to the correct asset, component and drawing? |
| Location fidelity | Can another competent person reproduce the measurement position? |
| Coverage adequacy | Does the inspected area resolve the expected damage pattern? |
| Measurement validity | Was the technique suitable and controlled for the material, geometry and condition? |
| Context | Are temperature, coating, surface, access and other relevant conditions recorded? |
| Traceability | Can the reported value be traced back to the procedure, equipment, personnel and raw or processed evidence? |
The complete FFS input package includes more than inspection data. For the broader design, operating, material and future-service inputs, see What Data Is Required for a Fitness-for-Service Assessment?.
Control Equipment Identity and Measurement Location
Every thickness value should be tied to a reproducible location system. Asset tag, component name, orientation and a clear reference coordinate are more useful than descriptions such as “north side” or “near the nozzle” when those descriptions cannot be reconstructed later.
The location system may use a marked-up drawing, isometric, vessel sketch, weld reference, clock position, elevation, distance from a datum or an encoded scanner coordinate. The exact convention is project-specific; consistency and reproducibility are the essential controls.
At minimum, the handover should make these relationships clear:
• Asset tag → correct equipment record.
• Component and feature → shell course, head, nozzle, branch, elbow, reducer or other defined item.
• Orientation and datum → a stable reference that can be used during repeat inspection.
• Reading or scan ID → a unique link between the marked-up location and the data file.
• Revision and date → the campaign and dataset version used by engineering.
A general NDT report should already identify the item, location, method and extent. NWE’s guide on how to read an NDT report provides that reporting context. For FFS, the additional question is whether the location detail is sufficient to define the damage model.
Design Mapping Coverage Around the Damage Question
There is no universal grid spacing or required number of thickness readings for FFS. Coverage must be capable of resolving the damage morphology that matters to the selected assessment route. A broad area of relatively uniform thinning and a narrow local pit do not require the same evidence pattern.
The inspection plan should therefore begin with the question that engineering needs to answer. Is the objective to confirm a general corrosion trend, locate the minimum wall, define the profile of a local metal-loss region, or compare a known area over time?
| Acquisition patternWhat it can supportMain limitation | ||
|---|---|---|
| Single spot reading | A targeted check at a known, reproducible location | Does not define the surrounding damage extent or prove the area minimum |
| Manual grid | A structured profile over a defined area | Suitability depends on grid design, location control and the scale of damage |
| Encoded corrosion map / C-scan | A spatially referenced thickness image over the scanned area | Resolution, near-surface capability, coupling and detection limits remain technique-specific |
| Targeted confirmation with another technique | Clarification of an anomaly or difficult geometry | Must be integrated into the same coordinate and revision system |
The lowest reported value is important, but it may not be sufficient. The engineer may also need the surrounding thickness profile, the length and width of the affected region, its relationship to welds or geometric discontinuities, and the boundary between damaged and unaffected material.
Method selection should follow the evidence objective, material, geometry and expected degradation. See NDT method selection for pressure equipment for the earlier decision about which technique is suitable; this article starts after that decision and asks whether the output is fit for engineering use.
Preserve Calibration, Procedure and Personnel Traceability
A reported thickness should be traceable to a controlled measurement process. ISO 16809:2025 specifies principles for ultrasonic thickness determination, but the applicable project procedure and edition must still be confirmed. The article does not reproduce the standard or replace the approved procedure.
The traceability record should identify the procedure and revision, instrument, transducer, relevant setup, material velocity basis, calibration or verification context, date, operator and personnel qualification scheme. The exact fields depend on the method and contract, but they must allow technical review of how the number was produced.
| Traceability elementWhy it matters | |
|---|---|
| Procedure and revision | Defines the controlled technique and the instructions that applied to the campaign |
| Instrument and transducer | Links the data to the configuration, range and technique actually used |
| Calibration / verification basis | Shows how the system was checked for the application and when the check applied |
| Material and velocity setting | Reduces the risk that an incorrect sound-velocity assumption biased the displayed value |
| Personnel and scheme | Shows who performed or reviewed the work and under which qualification system |
| Raw/processed data reference | Allows the reported table or map to be checked against the underlying evidence |
ISO 9712 specifies requirements for qualification and certification of industrial NDT personnel. SNT-TC-1A is different: it is a recommended practice used by employers to establish written practices for employer-based qualification and certification. The required scheme depends on the contract, jurisdiction and project specification; one should not be presented as a generic substitute for the other.
For a general explanation of the method and its applications, refer to NWE’s ultrasonic thickness testing guide.
Record Temperature, Coating, Geometry and Surface Condition
Field conditions can change the sound path, coupling or interpretation of a UT reading. These factors do not automatically invalidate the data, but they must be handled by the selected technique and recorded when they could affect the engineering decision.
| Context factorWhy it can matterWhat to document | ||
|---|---|---|
| Temperature | Sound velocity and transducer behaviour can change with temperature | Surface temperature, applicable correction or calibration approach, and limitations |
| Coating / lining | Conventional readings may include coating-related bias unless a suitable technique is used | Coating condition, measurement mode, removal areas or through-coating technique |
| Surface roughness / scale | Poor coupling or irregular interfaces can produce unstable or misleading echoes | Surface preparation, coupling condition, repeatability and rejected locations |
| Curvature / geometry | Probe contact and beam path may change on small radii or complex features | Component geometry, transducer choice, access and restricted areas |
| Material condition | Sound velocity, attenuation or structure may differ from the assumed basis | Material identification, velocity basis and any special verification |
| Access / obstruction | Uninspected areas may create a false impression of full coverage | Actual coverage, inaccessible zones and the effect on the damage question |
Manufacturer guidance confirms that sound velocity, roughness, curvature, calibration, coupling, temperature and operator technique can affect ultrasonic gauging. It also warns that coating can bias conventional measurements unless an appropriate method is used. The specific correction, range or capability remains instrument- and procedure-dependent.
Make Inspection Campaigns Comparable
Thickness trending is only credible when the compared values represent the same location and a sufficiently comparable measurement basis. A numerical difference may reflect degradation, but it may also reflect a shifted grid, changed coating condition, different technique, surface preparation, temperature or data-processing rule.
Before calculating or interpreting a trend, confirm that the campaigns share a stable coordinate reference and that changes are visible in a controlled change log. When direct comparability is weak, the dataset should be labelled accordingly rather than forcing a precise corrosion rate from incompatible points.
A campaign-comparability check should cover:
• Same asset, component and datum.
• Reproducible measurement locations or an explicit mapping between old and new coordinates.
• Comparable technique, probe configuration and coverage—or a documented technical justification for changes.
• Recorded differences in coating, surface preparation, temperature and access.
• Clear treatment of replaced, added, rejected or inaccessible points.
• Revision control that identifies which dataset was used in engineering.
Campaign data may later support life-extension planning, but the comparison should remain subordinate to a controlled integrity workflow. See the pressure vessel life-extension workflow for the broader stage-gate context.
Investigate Uncertainty, Outliers and Missing Readings
An outlier should not be deleted merely because it does not match nearby values. It may reflect poor coupling, a wrong echo, an incorrect location or a transcription problem. It may also indicate a real local minimum that requires confirmation.
The same principle applies to a no-read location. The cause may be surface condition, geometry, coating, equipment range, access or severe degradation. The disposition should be visible in the dataset and linked to a repeat measurement, alternative technique, expanded mapping or an explicit limitation.
| ObservationImmediate controlPossible next step | ||
|---|---|---|
| Unexpected low value | Repeat at the same controlled location; review waveform or measurement basis where available | Expand surrounding map or confirm with a suitable alternate technique |
| Unexpected high value | Check location, coating influence, echo selection and data entry | Re-measure and reconcile before trending or assessment |
| No reading / unstable reading | Record the location and field condition; do not enter a fabricated value | Change preparation, probe/technique or escalate the area |
| Inaccessible area | Mark actual coverage and the reason for exclusion | Evaluate whether the gap is material to the damage model |
| Conflicting campaigns | Check datum, method, context and revision history | Re-establish a controlled baseline if comparison is not defensible |
When the uncertainty can only be resolved in the field, a targeted in-service inspection scope may be more useful than attempting to model an unsupported geometry.
Build the Evidence Package an FFS Engineer Can Use
The engineering handover should allow the reviewer to move from the processed table or map back to the location, acquisition record and controlled procedure. It should also state what was not inspected and which assumptions remain.
A practical handover package may include:
• Equipment identification, component description and the applicable drawing or marked-up sketch.
• A controlled coordinate system, orientation, datum and unique reading or scan identifiers.
• Raw data or retained acquisition files where applicable, plus the processed table, profile or map used for engineering.
• Procedure and revision, instrument/transducer details, relevant calibration or verification records and personnel basis.
• Actual coverage, resolution or acquisition pattern, including inaccessible or rejected areas.
• Temperature, coating, surface condition, geometry and other context material to interpretation.
• Outlier, no-read and repeat-measurement dispositions.
• Comparison basis for earlier campaigns and a visible change log.
• Known limitations, uncertainty and questions that the inspection dataset cannot answer.
• Revision history and a single approved handover package used by the FFS engineer.
This is not a universal document list. The exact package depends on the damage mechanism, component, technique, assessment route and project specification. The objective is to preserve the chain from measurement to engineering input.
Measurement-to-Assessment Traceability Chain
| 1 | Asset tag and component |
|---|---|
| 2 | Drawing, orientation and coordinate |
| 3 | Reading / scan and raw evidence |
| 4 | Procedure, equipment, calibration and personnel |
| 5 | Processed map, table or profile |
| 6 | Engineering input and documented limitation |
| 7 | FFS assumption, result and follow-up action |
Decide Whether to Assess, Qualify or Reinspect
The correct next step depends on whether the dataset can answer the engineering question without hiding a material gap. “More data” is not automatically better; targeted evidence that resolves the governing uncertainty is more useful than a larger uncontrolled spreadsheet.
| Data conditionReadiness decisionRecommended route | ||
|---|---|---|
| Identity and location verified; coverage matches the damage question; raw and processed data are traceable | Ready for engineering review | Proceed to the selected FFS review. The assessor still confirms final adequacy. |
| Dataset is generally usable, but a context item or limited area remains uncertain | Use with explicit limitations | Document assumptions, evaluate sensitivity and define any monitoring or confirmation needed. |
| Spot readings do not define the extent or profile of local metal loss | Targeted reinspection required | Expand the map or use a suitable scan focused on the governing geometry. |
| Campaign locations cannot be reproduced | Trend not defensible | Re-establish the reference system before calculating a time-based trend. |
| Outlier or no-read has no technical disposition | Potentially material gap | Repeat, investigate and confirm with another suitable technique if necessary. |
| Procedure, calibration or qualification evidence is missing | Traceability limitation | Obtain the evidence or treat the dataset as limited; do not imply full validity. |
If the damage location or extent cannot be reproduced from the current dataset, close that evidence gap before relying on the data in an assessment. This may involve clarification, a focused repeat inspection, NDT audit or monitoring, or a defined FFS scope review—depending on what is missing.
NWE maintains separate service paths for NDT audit and monitoring, in-service inspection and Fitness-for-Service engineering. Share the available marked-up locations, raw or processed data, procedure and calibration context so the appropriate inspection or assessment scope can be defined without assuming that every case needs a full new survey.
Frequently Asked Questions
Is a spreadsheet of UT readings enough for an FFS assessment?
Only when the values are tied to the correct equipment and reproducible locations, the coverage is suitable for the damage question, the measurement process is traceable and the limitations are clear. A spreadsheet without those controls may be useful as a lead but not as a complete engineering basis.
How many thickness readings are required for FFS?
There is no universal number. The required quantity and spacing depend on the expected damage morphology, component geometry, technique, access, assessment route and project procedure. The objective is to define the governing damage, not to reach an arbitrary reading count.
Is the minimum thickness value enough?
Not always. Local metal-loss assessment may require the surrounding profile, length, width, orientation and relationship to welds or geometric discontinuities. The minimum value identifies a point; it does not automatically define the whole damaged area.
Can readings through coating be compared with bare-metal readings?
Only after the techniques and possible bias are understood and documented. Coating can affect conventional ultrasonic readings, and a through-coating or echo-to-echo method may behave differently. Direct comparability should not be assumed.
What should be recorded for each measurement location?
Record the asset and component, orientation or coordinate, reading or scan ID, date, method and relevant field context. The record should also link the value to the applicable procedure and controlled dataset.
How should outliers or no-read locations be handled?
Repeat and investigate them. Record whether the cause appears to be coupling, surface, coating, geometry, echo selection, access or possible severe degradation. Use an alternative technique or expanded mapping when the uncertainty is material to the decision.
Must UT personnel be certified to ISO 9712?
The required qualification or certification scheme depends on the contract, jurisdiction and project specification. ISO 9712 and SNT-TC-1A are different systems: ISO 9712 is a certification standard, while SNT-TC-1A supports employer-based written practices.
When is corrosion mapping preferable to spot UT?
Mapping is generally more useful when the decision depends on the spatial extent or profile of localized damage rather than a check at a known point. The appropriate scanner, coverage and resolution remain technique- and project-specific.
