Transformer Maintenance Work Packages: What Evidence Should Travel to the CMMS?

A CMMS work order is not the same thing as a diagnostic conclusion. The work order needs enough evidence for a planner, engineer, field crew, operations reviewer, and later auditor to understand why the work exists, what is in scope, what remains uncertain, and what happened after execution.

That is why the AI transformer maintenance work package should be treated as an evidence handoff problem. A bounded assistant can assemble and draft the package. The utility’s approved process still decides whether work is authorized, scheduled, dispatched, or closed.

What should travel with the work package

The package should preserve five linked layers:

1. Asset and scope. Asset ID, site, transformer rating, component, serial or equipment number, current service state, requested work type, and explicit exclusions. A DGA observation belongs to the correct oil system; an SFRA result belongs to a defined winding and test configuration.
2. Diagnostic evidence. Attach or link the original DGA report, online-monitor extract, oil-quality result, thermal or loading record, inspection note, PRPD record, SFRA trace, event record, and relevant maintenance history. Keep source, timestamp, unit, method, and data-quality status beside any normalized view.
3. Decision context. Include criticality, consequence of failure, spares, outage window, safety and environmental constraints, access needs, and the reason the proposed scope is proportionate to the evidence.
4. Approval state. Record the reviewer role, decision, rationale, conditions, expiry or due date, and escalation path. A draft recommendation must be distinguishable from an approved work instruction.
5. Execution and feedback. Reserve fields for as-found measurements, test-equipment identity, crew observations, photos, deviations, as-left results, parts used, and the engineer’s closeout decision.

CIGRE TB 962 organizes maintenance around planning, organization, execution, recording, and optimization. The work package should follow that lifecycle rather than stopping at “recommended action.”

Commissioning the evidence handoff

Before a new diagnostic integration goes live, test the mapping with a non-production record. Confirm that the source transformer ID maps to one CMMS asset, that timestamps retain their time zone or UTC convention, that units remain explicit, and that attachments are retrievable by authorized users. Test missing-data behavior: a blank sample date, unknown laboratory, stale monitor heartbeat, or duplicate work order should produce a visible exception.

For online DGA, compare monitor and laboratory records before allowing an alarm summary to populate work-package fields. For SFRA, retain instrument, lead placement, tap position, grounding, baseline, and comparison metadata; IEEE C57.149 emphasizes why those details matter to repeatable interpretation. For DGA, IEEE C57.104 is interpretation context, not a work-order generator.

The integration test should also prove the approval boundary. A draft can create a review task or a staging record. It should not silently write an approved maintenance status, dispatch a crew, change protection settings, or initiate an OT action. The exact control should be defined with the CMMS owner, operations, cybersecurity, and engineering authority.

What the field crew needs

The crew needs a compact, operationally useful package, not a model transcript. Include the work objective, isolation and permit references, equipment identity, approved test or inspection scope, acceptance criteria owned by the responsible engineer, hazards, drawings or procedures, and the evidence that motivated the work. Do not hide uncertainty: note whether the source is a laboratory measurement, online monitor, field observation, or illustrative planning assumption.

When the work is complete, capture as-found and as-left evidence in the same chain. If the crew finds no abnormality, that is useful outcome data. If a test is deferred, record why. If the scope changes, preserve who approved the change and when. The final record should make it possible to distinguish “condition confirmed,” “hypothesis not supported,” “evidence insufficient,” and “work could not be completed.”

Illustrative example, not a customer measurement

An illustrative work package for fictional asset TX-203 might show: an online gas trend flagged for review; a laboratory sample with a different collection date; a recent loading event; and a proposed repeat sample plus visual inspection. Those are not customer measurements. The package would list the source links, data-quality caveat, responsible engineer, and approval condition. It would not label a DGA threshold as an automatic maintenance order.

Security and information boundaries

Maintenance evidence can contain sensitive asset topology, operating context, outage plans, and protection or control references. NERC CIP-011-3 addresses protection of BES Cyber System Information; whether a particular record falls within scope is a utility compliance determination. Apply least privilege, approved retention, access logging, and a clear rule for which attachments may cross an integration boundary.

The sample evidence pack can help stakeholders agree on structure before connecting production data. The maintenance-window prioritizer adds planning context, while the audit-trail article explains why the decision record needs reviewer identity and outcome.

Product boundary

AgenticGrid Pro is described here as a bounded transformer APM workbench that can organize permitted diagnostic records, draft evidence summaries, prepare review questions, and stage CMMS fields. ProtectionAI is GridAPM’s separate Windows relay-testing application, useful for producing relay-test artifacts that may be attached to a work package. Neither product is the CMMS authority, an autonomous operator, a certification, or a substitute for qualified engineers and physical test equipment. Final decisions, maintenance dispatch, and OT actions remain engineer-approved.

References

• CIGRE. (2025). Guide for transformer maintenance (Technical Brochure 962). https://www.e-cigre.org/publications/detail/962-guide-for-transformer-maintenance.html
• IEEE. (2019). IEEE guide for the interpretation of gases generated in mineral oil-immersed transformers (IEEE Std C57.104-2019). https://standards.ieee.org/ieee/C57.104/7476/
• IEEE. (2024). IEEE guide for the application and interpretation of frequency response analysis for oil-immersed transformers (IEEE Std C57.149-2024). https://standards.ieee.org/ieee/C57.149/7354/
• International Electrotechnical Commission. (2024). Mineral insulating oils in electrical equipment—Supervision and maintenance guidance (IEC 60422:2024). https://webstore.iec.ch/en/publication/66421
• North American Electric Reliability Corporation. (n.d.). CIP-011-3: Cyber Security—Information Protection. Retrieved August 1, 2026, from https://prod.nerc.com/standards/reliability-standards/cip/cip-011-3
• U.S. Bureau of Reclamation. (2019). Reclamation strategic asset management plan. https://www.usbr.gov/assetmanagement/docs/Reclamation_Strategic_Asset_Management_Plan.pdf.pdf
• Tabassi, E. (2023). Artificial intelligence risk management framework (AI RMF 1.0) (NIST AI 100-1). National Institute of Standards and Technology. https://doi.org/10.6028/NIST.AI.100-1