Back to Main

Rules for Removing Water from Transformer Oil - made simple

reprinted from
NETA World
Spring 1999

Mark Lautenschlager, PE
President, ERC International, Inc.

When water enters a transformer some of it will be absorbed by the oil, some may be absorbed by the cellulose insulation, some may be trapped on the noninsulating structures in the transformer, and the rest of the water will go to the bottom.

Before attempting to filter or vacuum dehydrate the oil, verify the unacceptable water-in-oil result - it may not be truly representative of the water content of the transformer oil. Assuming that oil- testing laboratories utilize procedures that produce results within standard deviations, most inaccuracies are due to poor sampling procedures or "wet" oil in the sample valve. Always resample before attempting to dehydrate the oil. It is a good idea to also sample the oil from an upper valve and compare the results.

If the oil is determined to contain excess water, it is a good idea to perform percent power- factor/dissipation-factor tests on the transformer. These tests will determine if the cellulose has absorbed substantial quantities of water. It does little good to only dehydrate the oil if the cellulose is wet.

Free water in the bottom of a transformer does not necessarily mean that the oil or cellulose is wet. Good oil with high interfacial tension does not easily absorb water. It is a good idea to drain a little oil from a transformer after it has been deenergized and allowed to cool, particularly when the ambient temperature is just about freeze. By doing this, some of the water that was in solution in the oil at operating temperatures will be free water in the drained oil.. This is the least expensive way of partially dehydrating transformer oil.

Filter cool oil. Most filtering systems are designed to remove free water in the oil and only remove some of the water in solution with the oil. When the oil is cool less water will be in solution and more of the water will be free water; therefore, filtering is more effective. Standard filters can reduce water content to about 30 ppm. There are special filter systems that can effectively remove water content to about ten ppm,. but due to their slow filtering rate these systems are primarily designed to filter small quantities of oil or to continuously filter oil with the transformer in service.

Heat the oil when using vacuum dehydrating equipment. Vacuum dehydrating will remove water that is in solution with the oil and is more effective and more expensive that filtering. Vacuum dehydrating is necessary to reduce water content to less than 30 ppm, generally to five to ten ppm. In order to remove water when it is exposed to vacuum, the vapor pressure of the water must be raised. The only way to do this is to heat the oil. Vacuum dehydrating is not very effective unless the oil temperature is at least 100F, preferably higher.

1. Always verify elevated water-in-oil test results.
2. Always perform percent power-factor/dissipation-factor tests if the oil is "wet."
3. Drain some oil from cool transformers to remove some of the water.
4. When filtering oil to remove water, allow the oil to cool.
5. When using vacuum equipment to remove water, always heat the oil.

Mark Lautenschlager is President of ERC - Eleectrical Risk Consultants International, Inc., of Tampa, Fl. He is retired from High Voltage Maintenance Corporation as Vice President of Engineering, and is a Past President of NETA. He is presently a member of the NETA Standards Review Council.
Copyright © 1996-2010 NTT. All Rights Reserved.