I get a lot of questions about transformer oil analysis or Dissolved Gas Analysis (DGA) and what the results mean. So, here’s a quick overview on transformer oil analysis.
The insulating oil in a transformer is intended to serve two primary functions. First, as an electrical insulation to withstand the high voltages present inside the transformer. Secondly, to serve as a heat transfer medium to dissipate the heat generated within the transformer windings. The oil also helps extinguish arcs should a fault occur. Ideally the oil needs to maintain good electrical properties and resist thermal degradation and oxidation.
The life expectancy of a well-maintained distribution transformer generally ranges between 25 to 35 years. A Transformer operated under ANSI / IEEE loading conditions (ANSI C57.96) is said to have a normal life expectancy of 20 years. The life of a transformer, like most electrical equipment, is largely affected by maintenance, or lack thereof, and the environment. As the load is increased the heat generated, due to the winding and core losses, increases. Overloading transformers accelerates aging. As a “rule of thumb” for every 10-degree rise in operating temperature the life expectancy of the insulation system is reduced by one half. In addition to overloading, other factors such as overheating, harmonics, fault and lightning events further reduce the life expectancy.
Gases are formed in a transformer as a result of thermal and electrical disturbances, normal aging, and arcing temperatures. Regularly sampling and testing the oil for dissolved gases provides another tool for assessing transformer health. As with most electrical equipment maintenance testing, the overall trend is the best indicator of the equipment’s health. Studies performed over decades analyzing test results and comparing them to various failure modes have led to standards by IEEE, the Bureau of Reclamation, and others while research completed by Dornenburg, Strittmatter, Rogers and others has evolved into comparisons and ratios, based on empirical data, intended to provide insight into the internal condition of a transformer under operating conditions over time.
As part of the normal aging process insulation breaks down and internal conditions degrade. As insulating materials inside a transformer breaks down gases are produced. This occurs naturally and is to be expected during normal operation. If the gaseous byproducts produced are higher than expected due to normal aging, we can infer that a problem exists. Oil sampling is much like a blood test for a transformer. Like a blood test the results can give you an indication of otherwise hidden and potentially impending issues while being generally noninvasive.
Faults create mechanical stresses inside a transformer which affect the core, bracing, windings, and the connections. Faults can also lead to insulation degradation which increase the presence of specific gases in the oil. The Rogers ratio for example is a method for analyzing faults that have already occurred; as opposed to being a tool for detecting faults. The ratio provides an indication of what the problem might be, it doesn’t confirm you actually have a problem. It’s simply one more tool, which when effectively combined with insulation resistance, turns ratio, winding resistance, and power factor testing, helps ascertain the overall health and condition of the transformer.
The problem with interpretation comes into play since majority of the data collected and used in the standards has been based on the large power transformers used by utilities. Power transformers are found in high voltage transmission applications typically above 38kV and typically 200MVA and above. While distribution transformers are rated less than 200MVA and designed for end-user connectivity. The most common distribution transformers you’re likely to encounter are probably in the 38kV range with 480V secondaries. Along with size, capacity, and volume differences, the design of power and distribution transformer is different. In power transformers the I2r losses are minimized for a specific power flow. The design efficiency of a distribution transformer is in the 50-70% load range while power transformers operate at 100%. The magnetic operating point of the core is a design consideration in power transformers which changes the core mass, and the flux density is higher. These differences must be taken into consideration when interpreting gas results and, as with most maintenance testing, the trend is the critical piece.
In addition, the type of insulating fluid used, such as Mineral Oil, Silicone, or Natural Ester, have different standards and allowable levels change dramatically; and the result must be interpreted accordingly. Blindly applying the standards without a thorough understanding of the differences often contributes to misleading results. For example, while the C02/C0 ratio can be an indication of thermal decomposition. As a rule, if the individual concentrations are below 5000/500ppm, the results are thought to be less significant. When properly applied, a ratio below 3 may indicate ageing of insulation by arcing, while a ratio above 10 is indicative of cellulose ageing from thermal heating. Results should be viewed as part of composite analysis and are not individually indicative of a problem.
Moisture content, Interfacial Tension, and Dielectric strength, along with color, total gas, and the analysis of specific gases provide a picture of what may be occurring inside the transformer. Looking at CO and CO2 gases found in oil which originate from the oxidation of cellulose insulation over time can show if the transformer is aging naturally or if there is accelerated insulation break down occurring due to thermal decomposition. Combustible gases like hydrogen (H2), methane (CH4), acetylene (C2H2), ethylene (C2H4), and ethane (C2H6) are generated by thermal process and increases over time, and in the rate of change, can indicate potential problems are developing. Interpreting the results, requires an understanding of the testing procedures, loading and operating conditions of the transformer, the proper application of the standards and experience in analyzing the results. DGA becomes another maintenance test most plant and facilities professional do not have time to devote to developing the expertise in. By partnering with an experienced testing company, with expertise in this area, means you have one more tool in the testing arsenal and one less headache.
To summarize, for distribution transformers while Dissolved Gas Analysis (DGA) is a useful tool, the test is not conclusive, and the results are only useful to identify trends and provide insight into a transformer’s operating condition. Interpreting the results requires expertise to be understood and properly analyzed. In other words, DGA is simply one of the tools which, when combined with physical testing: Winding Resistance, TTR, Power Factor, Excitation Current, etc., provides data to develop a collective picture of transformer health.