Following this observation, corona inspection was done
during the clean fog periods at 3260 and 5000 h. In these
inspections, corona was detected in all the apparatus
insulation. On both bushings, corona was observed in the
region of the first few sheds from the energised terminal
where the simulation in Figs. 2a and b show the highest
electric field. In the surge arresters, corona occurred mostly
between the sheds along the housing. The simulations shown
in Figs. 2c and d indicate that the electric potential distribution
is similar along the surge arresters, but the electric field
strength on the surge arrester without pollution layer (sample
4) is higher because it has a shorter arcing distance (Table 3).
For this reason, it is probable that the higher electric field
strength caused the early degradation in sample 4.
At the end of the test, the four insulations showed
degradation along the surface as shown in Fig. 5. On the
bushing with pollutant layer, sample 1, a puncture in shed
no. 2 was observed (from the energised end). On the
bushing without pollutant, sample 2, erosion was evident on
shed no. 1 (from the energised end). On the surge arrester
with pollutant layer, sample 3, erosion was observed along
all the housing. The tracking damage that was observed on
sample 4 at 2500 h, Fig. 4a, changed very little after 5000 h
of testing, as evidenced in Fig. 5.
The obtained results showed that corona activity can
damage the polymeric insulation under dry and wet (clean
fog) conditions. Therefore the corona effect must also be
considered to evaluate ageing, in addition to the leakage
current pulses.