Managing PAUT & TOFD Scan Data in an Asset Digital Twin

Managing PAUT and TOFD data in a digital twin means attaching phased array ultrasonic testing and time-of-flight diffraction scan files, S-scans, and defect sizing results directly to the precise weld or component location on a 3D asset model, so the raw data, the interpreted result, and the physical location stay permanently linked rather than living in a separate file server.

Phased array UT (per ASME Section V Article 4 and AWS D1.1 Annex for welded structures) and TOFD have become the default advanced methods for weld inspection on new construction, in-service piping, and pressure vessel welds because they generate a permanent, re-interpretable data record rather than a single-frame radiograph or an operator's handwritten UT reading. That advantage is only realized, however, if the scan files remain accessible and correctly attributed to the exact weld years later. In practice, most inspection programs still store PAUT and TOFD files as raw data on a shared drive, named by job number, with no reliable link back to a specific weld ID on a specific asset once the original project folder structure is forgotten or reorganized.

Why Flat File Storage Fails Advanced NDT Programs

PAUT and TOFD data has more long-term value than a pass/fail radiograph because it can be re-analyzed as flaws grow, re-sized against ASME Section VIII Division 2 Part 5 or API 579 fitness-for-service criteria, and directly compared scan-to-scan to calculate crack growth rates. That value is lost when the data cannot be reliably located and matched to prior scans of the same weld. Common failure points include:

• Job-number-based folder naming that does not survive contractor turnover or project reorganization
• No standard linking between a weld ID on an isometric drawing and the corresponding scan file
• Raw data files (native PAUT/TOFD formats) separated from the interpreted report PDF, so re-analysis requires hunting down the original file separately
• No trending mechanism to compare flaw indications across inspection cycles at the same location
• Loss of encoder/positional metadata that would allow accurate re-registration of a follow-up scan against the baseline

Attaching Scan Data to the Asset Model

A digital twin resolves this by making the weld or component location, not the job folder, the organizing structure for the data. Each weld on a piping isometric or vessel model becomes a persistent node that accumulates scan history over time: baseline PAUT scan, subsequent TOFD monitoring scans, sizing results, and any resulting API 579 Level 2 or 3 fitness-for-service assessment all attach to that same node. When an inspector returns for a follow-up scan two years later, the system already knows the weld's inspection history, the prior flaw indications, and the exact positional reference used previously, making apples-to-apples comparison straightforward rather than a manual reconstruction exercise.

This is the model Atlantis NDT's digital twin platform uses for advanced NDT data: PAUT and TOFD files, along with RT and MT records, attach directly to the weld or TML location on the 3D asset, with full scan history retrievable by clicking the location rather than searching a file archive. For facilities running continuous monitoring programs on high-consequence welds, this location-anchored structure is what makes automated trend alerts possible — the system can flag when a new scan shows growth against the baseline without an engineer manually pulling both files.

Integrating with Plant Historian and CMMS Systems

Advanced NDT data rarely lives in isolation from the rest of a facility's operational data. Process conditions from a historian like OSIsoft PI, work order history from SAP or Maximo, and inspection scan data all describe the same physical asset from different angles, and integrity decisions are strongest when they can be viewed together — for example, correlating a TOFD-detected crack growth rate against a process upset logged in the historian around the same period. Atlantis NDT's guide on integrating a digital twin with SAP, Maximo, and OSIsoft PI covers the practical connection points for pulling process and maintenance data alongside NDT scan history into one asset view.

Justifying the Data Management Investment

Facilities weighing whether to formalize scan data management often ask what the return looks like in concrete terms. The clearest gains show up in three areas: reduced re-inspection scope when a follow-up scan can be precisely re-registered against a documented baseline, faster fitness-for-service turnaround because the sizing and location data is already structured rather than needing extraction from raw files, and lower risk of a missed re-inspection interval because monitoring welds are tracked as persistent assets rather than one-off job records. Programs evaluating the case for this kind of system can model the expected payback using Atlantis NDT's digital twin ROI calculator, which estimates savings based on fleet size, inspection frequency, and current data management overhead.

Practical Steps to Start

Facilities do not need to digitize every historical weld scan on day one. A practical rollout typically starts with the highest-consequence welds — those under active fitness-for-service monitoring, known crack-like flaws, or high-temperature hydrogen attack susceptibility — and establishes the weld-node structure for those locations first. Baseline scans are re-associated with the correct location, encoder metadata is preserved for future registration, and every subsequent PAUT or TOFD scan on that weld is required to attach to the same node rather than a new job folder. Expanding coverage from there follows the natural inspection cycle: as each weld comes up for its next scheduled advanced NDT scan, it gets folded into the structured system rather than back into a flat file archive. Over a few inspection cycles, the majority of a facility's high-consequence weld population ends up with a clean, location-anchored data history that supports both routine monitoring and the occasional deep fitness-for-service review without a data reconstruction project first.