Capacitive humidity sensors for transformer ageing monitoring

A transformer commissioned in 1966 was going to be scrapped due to age, high moisture content, and the need for increasing the capacity of the substation. To learn as much as possible about ageing condition from the unit it was decided to perform a scrapping analysis measuring the degree of polymerization (DP) of the paper, and the amount of methanol and water in the oil. In addition, three years before scrapping, the unit was equipped with sensors for continuously measuring water content and temperature in the oil during the last years of its service life.

The goal was to investigate how moisture dynamics and temperatures at the winding hotspots influence ageing as a test case for developing more advanced sensor-based digital twin models for assessing life consumption.

Using provided data of loading, top-oil temperature and water content, the thermal ageing of a power transformer was estimated. The calculated DP estimates (260-317) correspond well with the DP estimates from methanol (310-320) and with the measured DP values (257-301) from the scrapping study after the transformer was decommissioned. The results also show that water and oxygen were the main drivers of ageing. The hotspot temperature alone was too low (about 50 °C).

As long as the sensors are calibrated properly, using capacitive humidity sensors for continuous monitoring gives accurate water content which aids in improving the ageing estimation. Hence, making the use of such sensors feasible for developing sensor based digital twin models for online condition assessment.