Palladium alloy thin films measure hydrogen elegantly, with excellent parameters. But the whole industry faces one honest problem: short-cycle lab data is abundant, while real 10-year oil-immersed field data is scarce.
Datasheet specs are the entry ticket; 15-year total cost of ownership and long-term reliability are what decide. On that count, even CIGRE pours cold water: online DGA device life is typically only a quarter of the transformer’s.
Before any gas can be measured, it has to be driven out of the oil. Vacuum, headspace or membrane directly sets the error, the speed and the long-term reliability of the reading — and the difference is often overlooked.
A directly oil-immersed hydrogen sensor removes the degassing step — but not the conversion step. Turning activity into a ppm reading still depends on a coefficient tied to oil type and temperature.
Photoacoustic, NDIR, FTIR and TDLAS all fail physically on H2 — hydrogen has no infrared absorption. Yet almost every 9-gas online DGA claims to measure it. Here is where that hydrogen number actually comes from.
Propane and propylene appear in the standard that defines how to measure DGA gases — not in the standards that define how to interpret them. Here is what IEC 60599:2022 §5.8 really says about C3.
Datasheets are written to be compared badly. A weighted nine-dimension scorecard turns an online DGA monitor selection into a decision you can defend in a tender file.
Broadband infrared and narrow-linewidth laser sources are the two ways to build a photoacoustic DGA monitor. Here is what the choice changes — and what it does not.
The step-up transformer runs hotter, harder and closer to its limits than any other unit on site. That is why GSU partial discharge monitoring belongs on the main transformer.
A monitor that reports every hour but never raises an alarm adds data, not warning. Choosing the trigger condition is what decides whether condition-based maintenance for transformers pays off.
