While various maintenance strategies coexist, contingent upon the condition of the components and subcomponents of energy substation equipment, the reliability of power transformers remains the most crucial aspect from both technological and economic standpoints for distribution networks.
The failure of a power transformer can disrupt the energy supply to consumers and industries, leading to countless inconveniences based on the type of damage and the recovery time. This can result in revenue losses and diminished credibility, primarily due to the substantial costs associated with system outages, recurring failures and increased maintenance expenses.
To address the challenge of maintaining the operational longevity and reliability of power transformers, current maintenance strategies for substation assets employ a predictive maintenance approach based on prognostics. It’s important to bear in mind that prognostic tools do not evaluate the remaining operating time, instead, they can be utilized to assess future degradation, enabling a maintenance strategy focused on predictive maintenance.
In this context, the Dissolved Gas Analysis (DGA) of transformer oil emerges as the most informed and vital method to evaluate the health of a power transformer.
DGA analysis is the process of identifying and calculating the levels of gases that are formed within the oil immersed transformer operation, detecting incipient faults in the liquid or solid insulations where the aim is to support predictive maintenance of the power transformers.
As widely known, the two main causes of combustible and non-combustible gas formation inside an operating power transformer are electrical disturbances and thermal decomposition of paper and liquid insulation.
Considering that each transformer is manufactured individually and naturally generates gases to some extent at normal operating temperatures, the question arises: How can we determine whether the equipment is at risk or not?
The parameters that are normally used and most important to verify the health status of power transformers are dissolved gas analysis (DGA), oil quality analysis (OQA), and content of furfuraldehydes (FFA) in oil.
-
DGA: This is a method to measure the gas concentrations in transformer oil. When the transformer has faults different types of decomposition gases are formed, such as hydrogen (H2), ethylene (C2H4), acetylene (C2H2), methane (CH4) and ethane (C2H6) by the oil insulation decomposition, and carbon monoxide (CO) and carbon dioxide (CO2) when decomposition occurs in cellulosic insulation.
-
OQA: The combination of electrical, physical and chemical tests such as dielectric breakdown voltage (BDV), the water content, the power factor, the interfacial tension (IFT), acidity and color, each one with different weights, are used to prevent incipient failures and evaluate the replacement or recovery of transformer oil. The complete performed list of tests on the transformer oil can be seen in IEEE C57.106.
-
FFA: Its presence indicates the level of decomposition of the cellulosic material that constitutes the transformer solid insulation. Even just a small part of the furanic components is dissolved in the transformer oil, and there are no recommendations about the minimum value of FFA in the transformer oil on the standards, his content is a very important parameter of the life cycle of the equipment.
DGA analysis is a vital method in which the concentrations of various gases are associated with multiple techniques to interpret incipient faults in transformers. These techniques include key gases, Rogers, Duval Triangle, and Dornenberg methods.
Although the production of hydrocarbon gases from the oil doesn´t have a strong effect on your function as a coolant or electrical insulator, many international standards and brochures, for instance, IEC 60599 and IEEE C57.104, suggest action levels for key gas concentrations that range from an additional sample when normal limits have been exceeded, to a more extreme action such as taking the transformer out of service for further electrical testing and internal inspection when key gas concentration levels approach the action levels.
Let´s remember that after a sample has been taken and analyzed, the first action is to consider the concentration levels (in ppm) of each key gas in evaluating DGA results. But which key gas is most important and what happens if identify one or more gases in high concentration? What can we consider normal values?
Once the concentrations of key gases and their ratios exceed the recommended limits, it’s important to explore additional techniques to pinpoint potential issues within the transformer.
There are many doubts about normal gas concentrations in power transformer oil. Individual gas concentrations provide some information about transformer health, but is strongly recommended the evaluation of trends or rates for each key gas concentration for a better understanding and earlier detection of failure.
It is necessary to mention that ‘normal’ values must be determined through a statistical treatment of a large set of data from the transformer itself, where the normal concentration is that value considered in the range of 90% of the typical gas concentration values observed in the power transformer, as mentioned in the IEC 60599 standard. It means that each equipment has to be assessed individually comparing the actual values with historical values.
The rate of generation of each key gas depends largely on the temperature and volume of the material, oil, at that temperature. Comparing two transformer oils with different volumes, due to the volume effect, the equipment with the greater volume of insulation heated to a moderate temperature will produce the same amount of gas as the smaller one at a higher temperature.
Another observation that should be made when interpreting calculated trends is that when the time bases are longer, the trends will be more reliable. On the other hand, if the time bases are too long, trends increase more slowly, leading to later failure detection as a consequence.
This consequence is also related to the number of data points used, where the uncertainty of each point has a greater influence when fewer data points are used. To increase the number of data points simply increase the measurement frequency or time base.
As explained in the technical brochure CIGRE 771, the faults identified with hydrocarbon gases that seem to affect the paper insulation are considered more dangerous.
Basically, when cellulose insulation is oxidized due to thermal decomposition or electrical failures, the key gases methane (CH4), hydrogen (H2), carbon monoxide (CO) and carbon dioxide (CO2) increase.
The CO2/CO ratio is used to indicate the level of cellulose decomposition, where the ratio becomes significant when the gas concentration in the oil is above 5000 referring to CO2 and 500 ppm to CO. At this concentration, for a transformer with healthy cellulose insulation, the CO2/CO ratio must be between 3 and 11.
Extensively documented in the literature, a ratio below 3 indicates the occurrence of electrical arcing while a ratio above 11 indicates aging of the cellulose by thermal heating. This characteristic can be explained by the chemistry and design of power transformers where points with high temperature and low oxygen concentration, for example in the windings, will favor the production of CO, while low temperatures and high oxygen concentration, for example in regions with easy circulating oil – outside the windings, it will favor the production of CO2. As a consequence, a hot spot on the windings will demonstrate a higher CO concentration in the sample than one that occurred outside the windings.
It’s important to verify the oil temperature at which sampling is conducted, as the solubility of CO2 in oil significantly diminishes with rising temperatures. In normal conditions of operation, the rate of generation of CO2 is around 7 to 20 times higher than CO. The meaning of the CO2/CO ratio also depends on the historical values of both CO2 and CO (in IEC 60599).
With the deterioration of cellulose insulation caused by excessive heat, chemical substances are emitted, which, along with CO and CO2, dissolve into the oil. These chemical compounds are known as furanic compounds or furans.
In cases when the CO2/CO ratio is 3 or under with increased concentration of furans, is very probable that severe and rapid deterioration of cellulose is occurring and new considerations should be given to temporarily removing the transformer from service for a more comprehensive inspection (A concentration near 1500 ppb indicates a high risk of insulation failure). In transformers in good condition, there are typically no discernible furans present in the oil.
The generation of carbon oxide gases from paper insulation raises concerns about the deterioration of the insulation. Of particular concern is the charring of paper due to localized hot spots, especially within the windings, which can lead result in transformer failure. However, certain clarifications are necessary before translating diagnostics into actions:
-
Stabilization of the gas ratio after a significant change may not necessarily be a positive sign. It could indicate that the paper near a hot spot has been already consumed.
-
Even in cases of gas loss, since CO is lost to the atmosphere more rapidly than CO2, an increase in the CO/CO2 ratio can occur only if CO is genuinely being produced at a faster rate than CO2.
-
Certain changes in carbon oxide ratios are not necessarily indicative of faults. For instance, in transformers that have recently undergone degassing or immediately after factory heat run testing, the gas ratio may alter as gas trapped in oil-soaked paper insulation diffuses into the relatively gas-free bulk oil.
In summary, while CO and CO2 monitoring is valuable, it should be a part of a broader condition monitoring program that includes various techniques for a comprehensive assessment of transformer health. Regular maintenance and monitoring by trending the dissolved gas levels can significantly enhance the reliability of power transformers and reduce the risk of disruptions in the energy supply.