Age does not decide
An old transformer can be in good condition if it ran lightly loaded and was kept dry, and a relatively new one can be damaged. The decision rests on condition, on what can be corrected and on what a wrong call would cost.
What is evaluated
- Condition of the solid insulation: this is what cannot be restored without rewinding. It is inferred from dissolved gases, furanic compounds, moisture and electrical tests.
- Condition of the oil: if the problem is only degraded or wet oil, the oil can be reclaimed or replaced.
- Replaceable components: bushings, tap changer, pumps, fans, gaskets, controls and protections can usually be replaced.
- Mechanical integrity: short-circuit history and comparison of the frequency response with the reference.
- Load and failure history.
Business factors
- Criticality: what happens if the unit goes out of service, and whether there is a backup.
- Lead time: manufacturing a power transformer takes time; a repair or a spare unit can shorten the outage.
- Load growth: if the capacity is already insufficient, repair does not solve the underlying problem.
- Losses: a new design is usually more efficient; the energy savings go into the economic analysis.
- Life-cycle cost: repair, risk of failure, maintenance and losses, weighed against the investment in and expected life of the replacement.
Intermediate options
| Option | When it makes sense |
|---|---|
| Keep in service with monitoring | Acceptable condition and backup available |
| Treat or replace the oil and recondition accessories | Paper in good condition, problems in the fluid or components |
| Rewind or repair in a shop | Sound core and tank, localized winding damage |
| Replace | Severely degraded insulation, insufficient capacity or unacceptable risk |
When not to draw a conclusion
If the data are scarce or contradictory, it is better to extend the diagnosis than to decide. A useful rule is not to use any universal threshold: each decision depends on the unit, its function and the owner’s risk tolerance.