A significant and growing risk for unplanned outages and lost production has been increasing in the past decade due to failures of critical power transformers. With plant capacity running at all-time highs in the US, and little room for unplanned outages, it is more critical than ever that transformer reliability become front and center of any effective reliability program.
Oil testing can diagnose unseen problems inside a transformer. Mineral oil comes in contact with the internal workings of the transformer and contains valuable information regarding the condition of the transformer. Information about the equipment and a properly drawn sample are critical for making a correct diagnosis of the transformer's health.
Information about your electrical equipment is critical, but is of limited use by itself. Laboratory testing is also required to get the most accurate diagnosis, but accurate and representative test results are dependent on a good and representative sample.
Regarding containers for Karl Fischer Moisture Analysis, SDMyers has found that a glass bottle functions as an excellent container for moisture analysis, if all of the following conditions are met:
- Before sampling, the capped bottle is kept dry by using a desiccant tablet (discarded immediately before sampling).
- The glass bottle has a metal cap with a Teflon liner.
- When sampling, the bottle is filled to the very top (if possible), in order to minimize the gas space in the bottle.
- The KF Moisture test is the first test performed out of the bottle – i.e., the first time the bottle is opened after sampling is for the KF Moisture test.
Using proper sample containers will help ensure that your samples are representative of the actual conditions inside your electrical equipment, enabling proper diagnosis to extend the life of your transformer.
There are several methods for dehydrating oil in a transformer while simultaneously attempting to dry the paper insulation. However, drying the insulation (the most important step) is much more difficult than just dehydrating the oil.
Moisture may exist in a transformer in several forms.
1. Water vapor in the gas space of the transformer.
2. Water dissolved in the oil.
3. Free water suspended as droplets in the oil.
4. Emulsified water contained in the decay products of oil oxidation.
5. Water absorbed into and adsorbed onto the solid insulation.
6. Free water that has settled to the bottom of the transformer.
Paper has more affinity for moisture than oil does. Once wet, moisture removal is the corrective action that should follow. However, before moisture removal, it is important to first evaluate how the transformer became we.
The three most common ways to reduce the moisture levels in a wet transformer are:
1. Field Vacuum Dry-Out
2. Factory Dry-Out
3. Online Dryer
New transformers may have defects that can lead to failure. Frequently, such defects will leave signature dissolved gases in the oil. A timely Dissolved Gas Analysis may catch the fault as it begins, and before it advances far enough to do permanent damage.
Some defects may provide initial symptoms during the first ten months of transformer installations without causing or revealing more obvious indications until after the warranty period has expired. The timing of the first interval at ten months, and running the complete recommended package of tests, will serve to establish a diagnostic baseline.
A combination of dissolved gas levels and comments on the Rainbow Report provides DGA recommendations for new units. For newly installed transformers for of a small to medium size and class, SD Myers recommends retesting the dissolved gasses in three months, to obtain baseline data. For the larger, more expensive, and higher maintenance units, such as furnace transformers and generator step-ups, we recommend starting the DGA retest at one month.
Over the years, customers have asked us a variety of questions about dissolved gas analysis (DGA) results that indicate elevated ethane levels in natural ester fluid.
Natural ester-filled transformers, specifically those filled with FR3, have a tendency to generate ethane—and sometimes hydrogen—as stray gasses at temperatures normally found in a properly operated transformer.
While not typical in every case, it does happen frequently enough that we do not consider it to be abnormal.
As use of dissolved gas analysis (DGA) monitors increases as a growing component of transformer maintenance and reliability, it is imperative to understand the capabilities of monitors in their ability to align with conventional laboratory results and detect gas-related changes from a baseline. SDMyers studied DGA monitors from several manufacturers through experiments over 18 months. Technologies included in the study were gas chromatography, photo-acoustic spectroscopy, solid-state palladium, thermal conductivity detection, and selective membrane methods. This paper summarizes conclusions from that study based on technology employed.
Liquid power factor is an outstanding tool for evaluating in-service transformer oil. The test is valuable for acceptance testing of new oil from a supplier, and for evaluating conditions in newly installed equipment. For in-service oil, there are several adverse conditions that can be discovered from the liquid power factor results.
The values we obtain from liquid power tests and what these values indicate which condition, moisture, oil oxidation, or contamination, is causing a high power factor. The values we use to classify liquid power factor results for in-service oil are the same for all primary voltage classes of equipment.
For in-service insulating liquids, we recommend running liquid power factor at both 25°C and 100°C on all mineral oil-filled transformers except, perhaps, for very small distribution class transformers such as small pad mounted or pole mounted units. Liquid power factor is particularly important when D1816 dielectric breakdown voltage is also performed, since the results from the power factor may indicate the cause(s) of poor Dd1816 results. Equipment types where testing generally incorporates transformer testing packages, such as three phase regulators, should also be subjected to liquid power factor testing at both recommended temperatures.
How can you evaluate whether the solid insulation in your transformer is breaking down?
Furans form when the paper that makes up the solid insulation breaks down or depolymerizes. When this happens, the cellulose molecules break into shorter polymer chains and kick out a glucose monomer molecule. As this happens, the average polymer chain length in the paper (which can be measured and reported as the DP or degree of polymerization) decreases. Shorter polymer chains result in weaker paper, so that the tensile strength of the paper is also reduced.
Interpreting the results of furanic compounds can tell us whether the paper has broken down and, maybe even more importantly, whether it still is breaking down. SD Meyers tests for five different furanic compounds. Increases in the concentrations of these compounds indicate that the paper is breaking down. Further, the particular furanic compound(s) present provide some useful information on what conditions caused the paper to break down, and may even indicate whether those conditions are still present.
While knowing the state of the solid insulation is critical information for any transformer, it may not be the best use of resources to always run furans as a routine test. It is most important to analyze the insulating liquid for furanic compounds when you suspect a problem that the testing can help confirm and can help lead you to a solution. There are three very important times for performing analysis of insulating liquids for furans.
1. Test to obtain an initial baseline of furans for use as a diagnostic aid when abnormal conditions are later suspected.
2. Test when abnormal dissolved gas analysis results indicate that there may be a fault condition breaking down the solid insulation.
3. Retest furans at a shorter than normal monitoring interval to allow for closer monitoring of how furan levels are increasing and confirm how rapidly or under what conditions the solid insulation may be breaking down.
There are a couple of situations where performing the analysis on insulating liquid during every routine sampling and testing interval is appropriate.
The significance of the furan test results depends, at least in part on the reason for testing the unit in the first place. If you have baseline data or if you have other past history, any increase in total furans, and especially the presence of specific furans other than 2-furaldehyde, may be significant. Even a small increase may indicate a significant, ongoing breakdown in the solid insulation. When we do not have a history of furans analysis to follow up on other abnormal results or for a baseline determination, we have to make our judgement and recommendations based on complete information and some general guidelines.
Very high levels for furan test results may indicate substantial damage to the insulating system. The calculations that we perform to estimate the condition of the solid insulation by using the furans analysis results is performed in a two-step approach to estimate a DP and calculate insulation life remaining.
The key points of testing for furanic compounds are:
1. Why furan analysis gives us information on the condition of the solid insulation.
2. When do we recommend running the test?
3. What are the significant values?
4. How do we use the result of a furan test to estimate the condition of the insulation?
There are three methods for dielectric breakdown voltage testing. The oldest, the flat disk method is not very sensitive to the presence of moisture and not sensitive to changes in moisture content, unless the percent saturation of the oil is greater than 60%. The method is also not sensitive to the aging and oxidation of the oil. In spite of the fact that the application of this method may be limited in mineral oil filled transformers with regard to moisture increases and oil aging, D877 dielectric breakdown voltage determinations continue to yield important information for a wide variety of equipment types and insulating liquid types - including oil filled transformers. Therefore, SD Myers continues to recommend the test for purposes of providing that information.
The ASTM D1816 standard method of measuring dielectric breakdown voltage uses spherical VDE electrodes. This method is run at one of two gap setting: 1 millimeter or 2 millimeters. Because of the greater sensitivity, the rate of voltage rise is lower. Also, the D1816 test cell has a motor driven agitator that runs during the test to cause the oil to flow between the electrodes, carrying suspended particles into the gap between the VDE spheres where they can affect the breakdown voltage.
The D1816 dielectric breakdown voltage is a more sensitive method of dielectric breakdown voltage testing and is generally more useful than the D877 testing method under many circumstances. This method is sensitive to dissolved gases in the oil. As a result, high dissolved gas content in the insulating oil sample may depress the D1816 value to the point where it is outside the acceptable range. In short, oil that is acceptable in every respect that affects its performance in electrical equipment may still "fail" the D1816 determination because of dissolved gas content.
When D1816 test results are outside of the acceptable range, the first consideration is to cross reference the results to moisture, liquid power factor, and liquid screen tests to identify possible causes for depressed D1816 values.
D1816 Dielectric Breakdown Voltage Uses VDE Electrodes to evaluate new oil. The minimum D1816 values that are acceptable are 20 kV for the 1 mm gap setting and 36kV for the 2 mm gap setting. D1816 oil results that do not meet the minimum for the selected gap should be rejected.
The IEC method from Standard 60156 including values used in IEC standards for unused and in-service mineral oil uses electrodes that are similar geometrically to the VDE electrodes used in ASTM D1816. The spherical electrodes are spaced 2.5 mm apart, and the rate of voltage increase is 2,000 volts per second. The method in Standard 60156 allows the optional use of an impeller, operating in similar fashion to the one described for D1816, except that it operates at 250 to 300 rpm. The IEC method also allows use of a magnetic stirrer operating at a similar rate, if there is no significant chance that magnetic particles will be removed from the oil. The presence of magnetic particles would affect dielectric breakdown in the transformer, so removal of those particles by the stirrer during the analysis would yield unrepresentative values.
In Part 1 of a two part series on PCB testing, we will present some background information on PCBs, address when PCB testing is appropriate, and describe how a typical laboratory may perform PCB testing.
In Part two of the PCB series, PCB regulations, and the classification of electrical equipment based on PCB test results is discussed.
The Corrosive Sulfur test is a laboratory test performed on electrical insulating liquids of petroleum origin that detects the presence of corrosive sulfur in the sample. In Part 1 of this three-part series on the Corrosive Sulfur test, we will define corrosive sulfur, briefly review the history of the issue, and describe how corrosive sulfur can cause a transformer to fail.
When thinking about when the Corrosive Sulger test should be performed, it helps to consider the operating conditions of the transformers that have actually failed due to corrosive sulfer. Therefore, as a general guidline, if a mineral oil-filled transformer was manufactured or retrofilled since about the year 2000, and if it spends at least six months of the year with its top oil at a minimum of 70 degrees celcius, then we recommend that the oil be tested for corrosive sulfer.
Method A is the older method, which uses conditions of 19 hours at 140°C. Method B is the newer method, which uses conditions of 48 hours at 150°C.
Testing the oil in load tap changers provides valuable information on the operation of the unit. Monitoring conditions between preventive maintenance inspections is a critical step in preventing expensive maintenance problems and even unplanned outages. This first article in a series will address the aging of mineral insulating oil in an LTC.
You cannot just change the oil to lower the acid number to an acceptable level in place of a hot oil cleaning. Changing the oil will not permanently reduce the oxidation byproducts in the solid insulation. It is not an appropriate maintenance solution for lowering a high acid number.
Whenever there is excessive oxidation of the oil and aging of the paper insulation, the service recommended most often is hot oil cleaning.
This valuable process involves the use of a mobile oil processing unit (a.k.a. vacuum oil processor “VOP”) or oil reclamation “rig”, with the use of filtering clay such as Fuller’s Earth. The oil is reclaimed in the VOP through heating and vacuuming and sub-micronic filtration ─ and then returned to the equipment that is being cleaned. There are three steps in hot oil cleaning: (1) Thoroughly cleaning the oil of oxidation products; (2) Returning the heated, clean oil back to the equipment and performing circulations (i.e., repeating the cycle); (3) Monitoring the clay condition with oil testing/monitoring, making sure the process is doing its job.
Most mineral oil dielectric fluids contain an added oxidation inhibitor which is a chemical additive that acts as a preservative. The purpose of the inhibitor is to prevent oxygen from reacting with the oil, thus slowing the aging rate of the oil (and also of the solid insulation). The two most common oxidation inhibitors used in transformer oils are 2,6-ditertiarybutyl para-cresol (DBPC) and 2,6-ditertiary-butyl phenol (DBP).
Occasionally, we get questions from customers regarding whether two particular insulating liquids are compatible. For example, contemplating retrofilling with a different fluid, or adding fluid to a unit (“top-off”), but the original fluid is no longer being made. In any case, how do you determine a compatible substitute fluid?
As perchloroethylene fluid ages in service, it breaks down and forms hydrochloric acid. The AGE additive acts to neutralize this acid, so that the acid does not react with the metals in the transformers.
After the AGE test, the amount of acid remaining is measured, giving an indication of concentration of AGE in the sample. The result is compared to three calibration standards that are also run with the same procedure: a clank, a 1000 ppm AGE standard, and a 2000 ppm AGE standard. The response of the sample is compared with the response of the three standards to determine the concentration of AGE in the sample.
What causes bad D1816 dielectric breakdown voltage values? The first article in this series will discuss the three standard methods that SD Myers is equipped to perform, and why we perform them for our customers.
There are two standard methods from ASTM International: D877, Standard Test Method for Dielectric Breakdown Voltage of Insulating Liquids Using Disk Electrodes, and D1816, Standard Test Method for Dielectric Breakdown Voltage of Insulating Oils of Petroleum Origin Using VDE Electrodes.