Wednesday, 14 November 2012

Transformer Maintenance - Mineral Oils

Transformer maintenance used to be an all or nothing practices. Note that this is by no means a complete guide. It highlights some of the more critical parameters that a maintenance personnel should take note and hopefully gives you a sufficient overview of how complicated it is. A lot is to be considered in making an informed decision in regards to a transformer especially if it is a highly critical asset.

The best thing an asset owner can do to extend its life? Limit transformer operating temperature to 95 degree C.  Conservative estimates, every increase in 6 degree C halves the life of the insulation. In the field, the rule of thumb may be up to 10 degree C. Therefore, it is crucial to maintain the temperature instruments of transformer. Over the years, the instruments will be out of calibration. Some may be repaired, others may not. It is crucial to then compensate the annunciating point and hard & soft wired cooling system set point to continue operating under the safe set point. If left unchecked and uncompensated, it will lead to shorter transformer life and premature failure of the transformer.

Below is some of the critical parameters in DGA that one should take note. Oil sampling techniques and procedures are extremely important to get an accurate DGA result. Please have appropriately trained, experienced and skilled personnel for the task. I could not emphasize enough that it is a skill set by itself taking oil sample from transformer. A lot has to be considered and thought through from the materials of drain valves, oil collection techniques, temperature constraint of sampling window, humidity of environment, transformer oil tank pressure and etc. Surgical hygiene and precision is required to avoid minute contamination! We are talking about ppm, ppb level of contamination that is sufficient to cost you thousands of dollar of unnecessary down time and repair. Please utilise trained personnel for this task.

Transformer life is the life of the insulation. And the insulation degrades very rapidly with the presence of moisture and oxygen. Oxygen reading can be taken from DGA result. 3500ppm (volume ppm) or less should be brand new benchmark, 7000ppm should be a trigger to take action. Oxygen only comes from leaks and deteriorating insulation.

Insulation moisture of around 2% moisture by dry weight from DGA is the trigger point to begin degassing work and bring it back down to less than 1% at a cost of few cents per litre. It is far more cost effective to control moisture content from the outset. Average rate of water contamination in transformer with open-breathing conservators is around 0.2%/year. Membrane sealed conservator preservation system is around 0.03-0.06% per year. Water contamination will lead to swelling of insulation. Upon removal of moisture, the insulation will shrink back and loosening the clamp force of the core and coil. It is better to keep moisture level in check frequently than to suffer the consequences of repair. Note that water is distributed equally in the transformer when new. However, after some time of operation, it will congregate to cooler region, usually the lower one third of the insulation. Note down the sampling oil temperature (temperature at the bottom of transformer) and that should be able to assist in calculation of Moisture by Dry Weight to assist in decision making later. In the event that the lab does not provide M/DW percentage, refer to IEEE62-1995 - Myers Multiplier vs Temperature method. Then proceed to recheck with General Electric nomogram method which will give slightly higher reading. Do not make a dry out decision on a single DGA reading, it should be based on trends over a period of time. Once confirmed twice that W/DW is more consistently higher than 2% and oil is 30% saturated or more, arrange dry out as soon as possible.

Interfacial Tension should be determined along with the DGA. Good clean oil comes in on IFT number of 40-50 dynes per centimeter of travel of the wire. Oil needs to be urgently reclaimed at 25 dynes per centimeter, as sludge will start forming around 22 dynes per centimeter. Acid number also provides an indication of sludge formation. Acid number 0.4 is where sludging begins. Based on the data collected and published in AIEE transactions in 1955, you should hit the IFT limit 3-4 years before you reach that acid number anyway. The curve published seems to indicate the IFT number should cross with the acid number around the 0.2 mark making that the most cost effective point to carry out the reclaim. Useful to keep the acid number behind your mind in case IFT number is not indicating too much. Keep in mind acid buildup also accelerates insulation degradation and attacks the cellulose of the paper. Brand new oil should contain practically no acid as they are formed through the oxidation of insulation and oils as the transformer ages.

As mentioned before, oxygen inhibitor is a key item to extending life of transformers. Commonly used inhibitor is ditertiary butyl paracresol (DBPC). It acts as a sacrificial anode and oxygen would attack the inhibitor instead of the cellulose insulation. It will eventually be used up and should be tested for its presence. Ideal amount of DBPC is 0.3% by total weight of the oil. 0.1% reading is the trigger point to carry out re-inhibitation on the oil.

When cellulose insulation decomposes due to overheating, organic compounds are also formed. These chemical compounds are called furanic compounds or furans. Furans testing is to be included as part of DGA. It provides a reliable indicator for paper deterioration. Healthy transformer furans level should be non detectable or less than 100ppb. It is important to trend this reading. With thermally upgraded paper, Total Furans reading exceeds 1000ppb, the transformer has an estimated 40% life remaining. 1600ppb indicates a high risk of failure, where 2500ppb is essentially end of life. With non-thermally upgraded paper, 2200ppb of 2FAL is considered as excessive aging danger zone with 40% life remaining. 3800ppb carries a high risk of failure with 7300ppb is effectively end of life. Testing is completed for five different furans with different causes by different problems. The more common known causes are listed below:

  • 5H2F (5-hydroxymethyl-2-furaldehyde) caused by oxidation (aging and heating) of paper.
  • 2FOL (2-furfurol) caused by high moisture in the paper
  • 2FAL (2-furaldehyde) caused by overheating
  • 2ACF (2-acetylfuran) caused by lightning (rarely found in DGA)
  • 5M2F (5-methyl-2-furaldehyde) caused by local severe overheating (hotspot)

Along with the DGA, Dielectric Breakdown Voltage, 1mm gap, D1816 should be specified, and it should return a minimum of 28kV result. D877 is not as sensitive ti dissolved water and should not be used with oils for Extra-HV equipment. Dielectric breakdown tests do not replace specific tests for water content. (D1816, D877 and etc are all testing standards specified by ASTM)

DGA should be conducted regularly to confirm the trend. Any abnormal spike should be re-sample and tested to confirm the spike prior to determining cause and action. The above practices would allow the transformer to reach its designed life without too much issues (est 40 Years) provided the basic cleaning regime are covered and small corrective maintenance activities carried out without delay. Dynamic rating equipment for transformer is highly recommended.

Reference of this article:
  1. Transformers: Basics, Maintenance, and Diagnostics - Reclamation, US Department of the Interior Bureau of Reclamation, April 2005.
  2. Transformer Mid-Life Refurbishment - Prevention or Cure? - Kenneth J.Budin, Wilson Transformer Co. Pty. Ltd. TechCon 2001 paper.
  3. IEEE C57.12.90 - 1999
  4. IEEE 62 - 1995


  1. Transformer maintenance and electric motors in Australia is a very difficult task because the transformers are complicated devices and its parts and internal wiring is also complicated. Only experts can handle this.

  2. Hi David, thanks for the comment. Indeed it is a complicated equipment. It is not recommended to troubleshoot it yourself without an expert guiding you through. My view is, provided there's trained personnel to sample the oil, it is not recommended to touch it at all. Shall there be a trained personnel in hand for transformer oil sampling, the article's mentioned parameters are recommended for recording purposes and trending of transformer age and condition. This will enable accurate tracking of transformer ageing and schedule for replacement as they typically have very long lead time.