The problem of any dehydration of a power transformer is never only the removal of the water from its oil filling, but above all the safe removal of the water from its cellulose insulants (more than 95% of all water in the transformer is always deposited in cellulose materials).
Dehydration of new transformers – the safe removal of the water from cellulose insulants is not difficult in principle, but should be always and consecutively controlled e.g. by the TRACONAL to ensure:
the effectivity of a transformer dehydration - the only objective check of the dehydration - is the adequate correspondence between the measured decrease of the water content in the cellulose and the volume of the water discharged from the dehydrator
Over against – present checks of the transformer dehydration via water content in the oil ( the oil is sampled and the water content is measured immediately after the dehydration), are completely worthless due to subsequent dynamic migration of the water from the cellulose into the oil inventory.
Moreover any dehydration procedure must avoid:
The dehydration of ageing (or aged) transformers – is a lot more difficult due to potential brittleness of cellulose materials. The corresponding dehydration should avoid any other subsequent lost of its actual mechanical strength.
In this case a slow and a consecutively controlled on-line dehydration is obviously the best solution – See e.g. VS-06.