Abstract
Humidity can enter the insulation system when the
outer sheath of a cable becomes damaged. The
relative humidity can then increase above a critical
value (70 % RH), facilitating initiation and growth
of water trees. The purpose of this experimental and
numerical work is to determine how fast water
vapour will diffuse axially in a soaked cable with a
damaged outer sheath.
Before drying (evacuation) relative humidity and
temperature sensors were placed within the outer
sheath at different axial positions. After the drying
process a hole was cut at the cable end facilitating
water ingress. Numerical calculations of axial water
diffusion were performed using Comsol.
The numerical calculations show that the axial
water vapour diffusion in the cable is slow and
dependent of the air gap close to the swelling tape.
After 430 days the measured humidity at 0.5 m had
increased by about 75 % (initial 20%), and the
sensor at 1 m had increased by 40%. Numerical
calculations show a slower increase in the relative
humidity. The actual axial liquid water penetration
extensions, is yet not determined. The numerical
calculations show that this is an important factor, as
the calculations seem to be in more agreement with
measurements when adjusting the position of the
water front.
outer sheath of a cable becomes damaged. The
relative humidity can then increase above a critical
value (70 % RH), facilitating initiation and growth
of water trees. The purpose of this experimental and
numerical work is to determine how fast water
vapour will diffuse axially in a soaked cable with a
damaged outer sheath.
Before drying (evacuation) relative humidity and
temperature sensors were placed within the outer
sheath at different axial positions. After the drying
process a hole was cut at the cable end facilitating
water ingress. Numerical calculations of axial water
diffusion were performed using Comsol.
The numerical calculations show that the axial
water vapour diffusion in the cable is slow and
dependent of the air gap close to the swelling tape.
After 430 days the measured humidity at 0.5 m had
increased by about 75 % (initial 20%), and the
sensor at 1 m had increased by 40%. Numerical
calculations show a slower increase in the relative
humidity. The actual axial liquid water penetration
extensions, is yet not determined. The numerical
calculations show that this is an important factor, as
the calculations seem to be in more agreement with
measurements when adjusting the position of the
water front.