Visit Count: 583 DGA-Assessment of measuring accuracy. As all of us know the DGA test for transformer oil, is an important test for the assessment of the transformer health. There are a lot of papers and methods guiding to do the thing in the right way. Rogers Ratio method, Duval Triangle Method, Key Gas Method are some of the leading names. However, the aim of today's article is not about discussing these methods but here I will put some fundamentals which will be useful for understanding the basics only. First of all, whenever we call this method a Dissolved Gas Analysis, that means it is not like that some gas bubbles get formed in oil due to some heavy spark inside the transformer, and that gets dissolved into the oil. It may happen but it is a very extreme case when Buchholz operates and in that case, there remains nothing like Dissolved Gas Analysis. Instead, first, understand that the transformer oil is a hydrocarbon compound and hydrocarbon molecules are in the form of a chain containing 50-60 carbon atoms in each chain as shown in the following fig. Whenever due to any reason, say copper heating, insulator heating, or arc inside the oil; some parts of these hydrocarbon chains breaks and remain in oil as gas molecules. when the percentage of these gas molecules exceeds a certain limit (saturation limit of that gas for the oil) it appears in Buchloz relay in the form of gas. The following figure illustrates the generation of gas molecules due to the dissipation of energy inside the oil. The Takeaway: Detecting Gasses (Abnormal) in Oil (Dissolved Gas Analysis) is sure and very early detection of a problem inside the transformer. Moreover, this method does not require any outages or heavy machinery. Previously this can be done only in sophisticated labs, but now with the advancement of the new techniques, we can do such type of testing in fields too. While performing DGA on the field the most important care to be taken is the accuracy of the measurement. All analysis fails if accuracy is not maintained. Here the problem is the calibration of the DGA kit. The manufacturer of the existing kit claims that there is no need for its calibration and the kits are self-calibrating. But whenever important decisions needed to be taken depending upon these results it is foremost important to ascertain the accuracy of the equipment. Ascertain accuracy by mixing the new oil with used one To ascertain the accuracy of DGA kit Er. Vasant Sonawane has performed some simple experiments which are put here. In the experiments, first two samples of oil were taken. One sample having a substantial quantity of dissolved gasses and the other sample having a very less quantity of dissolved gasses. Then these two samples mixed in fix proportion of 15 ml good oil and 35 ml contaminated oil 25 lm good oil and 25 ml contaminated oil 35 ml good oil and 15 ml contaminated oil Again DGA results of these mixed oils were taken and compared with theoretical values. Following is the table (click to download and enlarge). The measured and theoretical results are almost tallying, though there is some scope for improvement. Comparison of results from two different kits Also, a comparison of DGA results by two different kits was carried out and the results are as below. from the result, we can easily conclude that the kit bearing Sr. No. 80-0905 (Parli) failed to detect the Hydrogen content. From a diagnostics point, proper detection of Hydrogen Gas is crucial because it is the gas that is common for almost all faults. The Takeaway: If in doubt, you may compare results by two kits. They shall match nearly, as required for a practical purpose. Comparison of results for two samples by same kit Here is the comparison of results for two different samples of the same oil by one kit is done, and the results are as below. The Takeaway: Generally it is the practice to collect oil from the top and bottom of the transformer. When the transformer is in service oil is continuously circulating by the thermosiphon effect. Hence there is no reason that the DGA results of these results may vary. Hence in the case of a transformer in service, the method of testing two samples may be viewed as a confirmatory test for the DGA results when there is not much deviation in the test results. Conclusion When sufficient care and efforts are taken, DGA in the field is a very good tool for the health monitoring of the transformer.