Fleet Monitoring Is a Portfolio Problem
For a utility, DGA monitoring is not about any single transformer — it is about a portfolio of assets that spans voltages, vintages and criticalities. A fleet manager faces a different question from an owner of one flagship unit: how do I get a usable condition signal across hundreds of transformers without spending like each one is a critical asset?
The pattern emerging in utility procurement is telling: online dissolved gas analysis is increasingly bought at fleet scale, as a standard control-system component rather than a special project. The specification is no longer “which monitor?” but “which monitoring layer, at what cost per asset?”
A Two-Layer Architecture
The cost-efficient answer is to separate screening from confirmation. A hydrogen sensor on every transformer provides the continuous screening layer; a full multi-gas system is then placed only where fault risk or asset criticality justifies it.
| Layer | Coverage | Monitor | Signal |
|---|---|---|---|
| Fleet screening | All units | DGA-500 hydrogen monitor | H2 level + rate of rise |
| Critical confirmation | Largest / most loaded units | DGA-900 9-gas + moisture | Full fault typing per IEC 60599 |
| Event escalation | Alarmed units | Laboratory DGA | Condition rating and repair decision |
This mirrors the online DGA selection framework: screening first, confirmation where it earns its cost.
What Fleet Data Must Do
A fleet program only works if the data lands somewhere that drives action. Four requirements decide whether the investment pays:
- Single control-room view: MODBUS, IEC 61850 or MQTT so every unit reaches the same SCADA or cloud view — see communication protocols.
- Per-unit baselines: each transformer’s own normal level captured at commissioning, so alarms are set on deviation rather than a generic threshold.
- Rate-of-rise alerts: the rate of change distinguishes a drifting baseline from a genuine upward trend.
- Remote operations: unmanned substations need remote monitoring and IIoT access, not a site visit to read a panel.
Procurement and Specification
When the monitoring layer is bought at fleet scale, the specification needs to be repeatable. Essential items include certified performance (for example CEPRI A1 or equivalent evaluation), no consumables, a defined calibration interval, and a total-cost-of-ownership model rather than first price. For a full checklist, see the DGA monitor RFQ technical specification.
Zero-consumable, low-maintenance sensors are the practical enabler of fleet scale: a unit that needs a carrier gas or periodic cell replacement is not something a utility wants to manage across hundreds of sites. A phased roll-out helps too — pilot on a representative sample of units first, validate the data path and alarm handling, then expand to the full fleet once the operating model is proven. The 10-year total-cost-of-ownership comparison shows how the per-asset math changes when consumables and service visits are removed.
From Screening to Fault Typing
Hydrogen screening answers “is something starting?”; it does not answer “what type of fault?”. When a screening alarm escalates, the unit moves up the tier: a 9-gas system or a laboratory sample types the fault per IEC 60599 / IEEE C57.104, and the repair decision is made on evidence rather than guesswork. For the reasoning behind starting with hydrogen, see hydrogen vs multi-gas DGA.
PAS DGA for Utility Transformer Fleets
PAS DGA designs the monitoring layers that make fleet-scale DGA practical: the DGA-500 hydrogen monitor for portfolio-wide screening and the DGA-900 for critical-asset confirmation. With the 2026 regulatory push toward mandatory online monitoring covered in the compliance guide, the fleet-scale question is increasingly “how quickly”, not “whether”.
Contact PAS DGA for a fleet monitoring plan matched to your portfolio size and voltage mix.