A power system is a complex network of key high-voltage components working in synchrony to provide electricity. Power transformers, which help step up and step down voltage so power can be transmitted over long distances, are designed to be in continuous operation. Hence, the life and reliability of transformers is of paramount importance.
Multiple factors affect the life and performance of a transformer. Stresses including higher voltage spikes, fault current, harmonics, overloading, and transportation can deteriorate the transformer’s performance characteristics. Regularly scheduled maintenance can help detect these problems at an earlier stage and improve the life, reliability, and performance of the transformer.
Transformers are available in several types and all shapes and sizes. If we break it down to the bare minimum components required to form a transformer, transformers are simple apparatuses. Two coils of insulated wire, commonly referred to as windings, are wound in proximity to an iron core. One winding is connected to a power source, such as the utility or a generator, and is designated as the primary. The other winding provides power, which is transformed in some manner, to the load and is designated as the secondary.
TRANSFORMER TESTING
All transformers have three fundamental components: insulation, winding, and a core. All electrical tests are designed to detect problems within the integrity of the winding structure, the insulation materials, and the core. Tests that can be performed include the transformer turns ratio (TTR) test, winding resistance (WR), and short circuit impedance (Z% or SC impedance). These tests provide insight into the integrity of the winding structure and the core. Insulation power factor (PF) and oil testing, namely dissolved-gas analysis (DGA), provide data indicative of the transformer’s insulation health. Combining these tests gives a complete picture of the transformer’s health.
TRANSFORMER TURNS RATIO (TTR) TEST
Over time, advancements have reduced testing time and increased our understanding of the voltage needed for testing. This article discusses how to leverage physics to amplify accuracy while reducing test time by injecting voltage from the low-voltage side with a much smaller footprint. One of the quickest and most popular tests for transformer performance characteristics is the transformer turns ratio (TTR) test, which has become a go-to test for transformers of all sizes.
Conventional Turns Ratio (TTR) Testing Methodology
One reason the TTR test is popular is the fact that once the transformer is offline, the TTR test setup is the most simplistic and the test runs on each phase in less than 30 seconds. The TTR test’s simple but powerful detection can quickly identify any open windings or any turn-to-turn shorts. This becomes the characteristic test for a transformer.
According to IEEE Std. 57.12.90, Standard Test Code for Liquid Immersed Distribution, Power, and Regulating Transformers, the turns ratio for any transformer is defined by the ratio of the physical number of turns on the high-voltage winding to that of low-voltage windings. TTR is performed at rated or lower voltage and rated or higher frequency. If the transformer has taps, the turns ratio is determined for all taps for the winding. In the case of three-phase transformers where each phase is independently accessible, single-phase testing can be performed, although three-phase testing, using three-phase TTR test equipment, can be performed more efficiently.
The tolerance for IEEE, NETA, and other European and Australian standards for TTR has universally been 0.5% for liquid-filled or dry transformers. Let’s take an example for a single-phase transformer rated 7,200 V to 120 V; the nameplate (NP) turns ratio is 60:1. Given a maximum of 0.5% deviation, a reading below 59.7:1 or any reading higher than 60.3:1 would be considered a failed result. TTR values between 59.7 and 60.3 indicate the transformer has passed the various tests.
The only exception is for transformers with tap changers, specifically on-load tap changers (OLTC). The limits apply to the highest and the lowest taps along with the neutral (nominal) position. This is because OLTC designs can sometimes have quarter turns from where the manufacturers can pull out leads for those tap connections and may exceed the limits when a manufacturer has a +/-10% regulation in 32 steps, making each step a 0.625% increment or decrement in voltage from the nominal voltage.
Now that we understand what TTR is with respect to the transformer, we can discuss the factors impacting the actual TTR values when using any test instrument.
With a 138 kV rated primary voltage and a 4.365 kV rated secondary, the best way to do a TTR would be to inject 138 kV into the primary and measure 4.365 kV on the secondary using a voltmeter. However, voltmeters are not built for these high-voltage ranges. Therefore, potential transformers (PTs) are required to step down the voltage to read within the limits of the voltmeter. However, doing so adds to the overall inaccuracy in the readings. Additionally, this test requires a large transformer to achieve the voltage ranges for the transformer under test. For this reason, testing at line voltage becomes impractical.
For TTR, an assumption is typically made that TTR = NP/NS ≈ VP/VS (Figure 1). This assumption means that the excitation current has a negligible effect on the value of the turns ratio, which is true. However, the influence is not zero.
Excitation current is the minimum amount of energy absorbed from the primary side by the core to start acting as a transformer to output voltage on the secondary. Excitation current is responsible for a very slight change in magnitude of Vs = (Ns/Np) * Vp and a slight deviation in phase angle. The amount of change is dependent on how large the value of the excitation current is for that voltage. It is easier to understand excitation current as a minimum amount of fuel consumption when idling when looking at mileage per gallon in driving for any vehicle.
If a transformer is rated for 7,200 V on the high side and 120 V on the low side, with a TTR of 60:1, the turns ratio is going to be linear. Hence if we inject 7,200 V on the high side, we will get 120 V on the low side — or even if we inject 60 V on the high side — we will get 1 V on the low side. Then what is the correct voltage to inject on any transformer and is this dependent on the transformer’s rated voltage?
As the test voltage is increased, the excitation current increases non-linearly and becomes a smaller percentage of overall energy supplied by the test set, resulting in increased test accuracy. Figure 2 is an example of the variance that can be obtained on a large power transformer with 138 kV high voltage on the primary and 4.365 kV low voltage on the secondary side. The testing was done by single-phase injection on the high-voltage side and measuring on the low-voltage side. It is evident that with lower voltages such as 8 V when performing TTR on larger transformers with TTR values greater than 25, the variance can be high enough to fail the standard on a perfectly good transformer with a near-zero error.
As the test voltage increases, there is a rapid increase in accuracy initially but after a certain voltage, we have laws of diminishing returns on the accuracy of TTR. The higher the value of TTR, the better it is to use a higher voltage while measuring it.
Step-Up Methodology
Unlike step-down methodology where voltage is applied on the high side and measured on the low side, step-up methodology inverts this: voltage is applied on the low side and measured on the high-voltage side (Figure 3).
Here, we are using the transformer geometry to our advantage for stepping up the voltage and then measuring it. For example, if the TTR value is 100:1 (or 100), we can apply 2.5 V on the low-voltage side to read 250 V on the high-voltage side, and with a 2.5-V injection, we can get a reading as if the voltage injection was 250 V.
One major challenge to this approach is that if we were to apply 250 V on the low-voltage side, we’d get 25 kV on the high side, which can be dangerous for the test instrument and the personnel testing it. To avoid this, we recommend a very quick voltage injection from the high to low side at a lower voltage to gauge the TTR value with a rough approximation, and then record the results when injecting from the low-voltage side.
The hardware is the other challenge. A higher-voltage supply that can handle a very low current value has typically been used. As we are injecting from the low side, we need a generator that is lower voltage but can support higher values of current (typically up to 2 A, instead of 0.5 A when using a higher-voltage supply). Development in modern electronics has brought in very capable (higher VA) power supplies or inverters with 24 V and 48 V architecture in much smaller footprints. We used this size advancement to use a 48 V, 2 A generator instead of a 250 V, 0.5 A generator with equivalent VA of 96 VA and 125 VA. While using the lower voltage as we inject from the low-voltage side, we gain better accuracy if the injection is from the high-voltage side.
Now that we’re supplying voltage from the low side, the excitation current is higher with respect to the lower voltage on the secondary. However, the equivalent excitation current from the higher voltage is the same as if the equivalent were pushed from the higher voltage side of the transformer. The main cost and size preventative challenge is resolution. The better the resolution, the greater the accuracy of the TTR instrument. Usually, the supply voltage can be measured at a higher resolution. By supplying on the low-voltage side, we get a higher voltage on the high side; hence we see a 0.1 V resolution. It’s better to measure 100 V on the measuring side by applying 10 V on the supply side, as we have an error of 99.9 V to 100.1 V. By contrast, if we were to measure only 10 V, the result would be a tolerance of 9.9 V to 10.1 V.
This makes measuring the voltage on the low side more accurate with a voltmeter of the same resolution.
CASE STUDY
This case study uses a single-phase 449-MVA autotransformer with a primary nameplate voltage of 228 kV and a tertiary nameplate ratio of 26.4 kV. An exceptionally large transformer was considered in this case because it has historically been difficult to test larger transformers, especially those with a tertiary winding, due to the much higher voltages required for the step-down methodology. As can be seen in Figure 4, if the testing were done only up to 100 V using the conventional method, the maximum deviation on Tap 1 is 0.51%, which is just outside the 0.5% limit set by NETA and IEEE for this transformer.
The next step would be to investigate whether there are any turn-to-turn shorts on the low side, as the measured value is higher than the calculated value, which would indicate a turn-to-turn short on the low-voltage side. However, the testing done at only 20 V from the low-voltage side provides much more realistic TTR values. This proves the step-up method superior given a transformer with a maximum of 0.32% deviation at Tap 1 compared to 0.51% when voltage was applied on the primary.
CONVENTIONAL SHORT CIRCUIT IMPEDANCE TESTING
The short circuit (SC) impedance test plays a crucial role in monitoring the overall losses incurred at full load as well as short circuit calculations. While SC impedance (Z%) is always performed at the factory, it is not often tested in the field. The use of SC testing has not been prevalent even though it offers increased value as an early indicator of a developing issue. One reason has been the necessity to short the secondary side using copper bars or cables with a larger cross-section to handle the higher current flow during testing. Results analysis involves comparing the measured SC impedance results with the nameplate data to determine the percentage of deviation from the rated values.
The lack of standardized criteria for selecting the appropriate size shorting cables for different transformer types complicates understanding the results. This article explores recent advancements in technology that have helped streamline the shorting process on the secondary side, simplifying the short circuit testing procedure with a consistent short on any transformer, and improving results analysis due to consistent testing.
Short circuit impedance value is a critical power transformer specification that is guaranteed by the manufacturer. The manufactured transformer in the factory will be tested, and the measured short circuit impedance value must be equal to the calculated value in the design stage with an allowable tolerance. It is also considered a routine test but is rarely done due to the nature and rigorous preparation required before conducting it.
The importance of the short circuit impedance becomes apparent when the short-circuit test results do not match the guaranteed values. This error causes problems such as a lack of conformity with standards, increased load losses, and increased voltage drop at the in-service transformers. These design parameter values impact financial implications related to the transformer operation. When short circuit impedance is on the higher side, the maximum short circuit current, voltage regulation, and efficiency are reduced.
To simplify, the percentage impedance of a transformer is the percentage of the rated voltage applied at one side (primary winding) to circulate rated current on the transformer keeping its other side (secondary winding) under short circuit conditions. The current is called a short circuit current, which is high magnitude due to the short circuit on the secondary winding.
The primary voltage is measured right at the transformer’s primary terminals to avoid errors due to the voltage drops across the ammeter and wattmeter. The input voltage is increased from zero until the ammeter in the primary circuit indicates normal full-load primary current. When this occurs, the normal full-load secondary current is circulating in the secondary winding. Because the secondary terminals are short-circuited, the input voltage required to produce full-load primary and secondary currents is around 3% of the normal input voltage level. With such a low input voltage level, the core losses are so small that they can be neglected. However, the windings are carrying normal full-load current, so the input is supplying the normal full-load copper losses and the output power is zero, which also helps the wattmeter to measure true full-load copper losses.
Procedure and Calculations
The transformer to be tested must be offline with no connections on the bushings. Bushings need to be free of dust and moisture. Another critical step is selecting the shorting leads used to short circuit the secondary side windings. Considering that full-load current must flow in the transformer to calculate the impedance, it is mandatory and important to use very thick shorting leads or, even better, use copper shorting bars. If small and thin shorting leads are used for testing, there is a high possibility they can break due to the large magnitude current flowing through them. This creates a safety issue for operating personnel around the test zone.
The contacts should also be clean and tight, and the shorting leads must be as short as possible. There are no specifications for the size of the copper wire to use as shorting leads depending on the size of the transformer. In our testing experience, the recommendation for a distribution transformer is to use wire greater than 1 AWG. As transformer size progresses, copper or aluminum bars are to be used for shorting purposes.
After connections are made, the testing is started by applying minimum voltage on the primary side. The minimum voltage also keeps the flux density of the core to a minimum, reducing the iron losses to almost zero. The impedance calculation is the summation of resistance and reactance referred to the primary side. The wattmeter measures W (full-load losses) and the ammeter measures (Isc) short circuit current. Phase angle is also measured depending on the test equipment being used to run the test.
Challenges with Impedance Testing
There are three common hurdles while conducting short circuit impedance testing.
- The first hurdle is the test equipment used to run the test. Test equipment must be calibrated before running the test because the test parameters are measured and calculated to provide the results. The test operator needs to be well-trained to operate and run the test using that specific test equipment.
- The second hurdle is testing a single phase against three phases for a three-phase transformer. In this case, the test operator would have to run the test on all three phases individually, which involves switching the test leads on all three phases and calculating the three-phase equivalent short circuit impedance, which can give erroneous results.
- The third frequent hurdle is the quality of shorting leads as discussed in the previous section. The quality of shorting leads is responsible for the circulation of the full-load current through the primary as well as a major safety concern while running the test (Figure 6 and Figure 7).
Improvements to Short Circuit Impedance Testing
Recent developments in technology have enabled the testing industry to run the short circuit impedance test smoothly without the hurdles discussed in previous sections that usually affect the testing procedure and results. This technology uses the feature of internal shorting using high-rated resistors in the test equipment instead of using external shorting leads. In this case, the internal shorts are connected on the primary side. The losses in the internal shorts are also measured and eliminated from the calculations to provide accurate results. The test equipment and the test connections are the same as for the TTR and winding resistance test (Figure 8).
As shown in Figure 8, the secondary side is the supply side and the primary side is short-circuited. As the test progresses after the minimum voltage is applied, excitation voltage, current, phase angle, and losses in the internal short are measured. Once the testing is done, short circuit impedance values are calculated and compared to the nameplate ratings of the transformer, and the percentage difference is calculated to verify whether the difference is under the limits as per the IEEE standards.
CONCLUSION
The evolution of technology has resulted in advances in the landscape of testing. From the early days of rudimentary methods to the sophisticated, technology-driven approaches of today, testing has become more nuanced, efficient, and effective. This evolution drives advances to make testing faster and more compact, while at the same time, being more accurate and precise.
The continued progress in understanding the basics of testing instruments as well as the assets we test helps us make testing faster, smaller, and more accurate over time. For example, for TTR testing, we can take advantage of the transformer’s configuration to use a smaller test instrument, reduce the test voltage produced by the test set, and increase the accuracy of the measurement.
In the case of short circuit impedance, overcoming the hurdle of having copper bars as shorting on the secondary by using the test instrument to short the leads internally gives a consistent short each time. This improves the repeatability of the test and makes the test more accessible for those who do not have the right size copper bars for testing in the field. It also reduces the test time, making it more practical in the field rather than just being a reliable factory/lab test.
REFERENCES
[1] IEEE Std. C57.12.90–2021, Standard Test Code for Liquid Immersed Distribution, Power, and Regulating Transformers. [2] ANSI/NETA MTS–2023, Standard for Maintenance Testing Specifications for Electrical Power Equipment and Systems. [3] A. Teymouri and B. Vahidi. “Power Transformers Short Circuit Impedance Calculation Using Energy Method Based on Leakage Flux,” 2023 3rd International Conference on Electrical Machines and Drives (ICEMD), Tehran, Iran, Islamic Republic of, 2023, pp. 1–7, doi: 10.1109/ICEMD60816.2023.10429519. [4] IEEE Std. C57.152–2013, IEEE Guide for Diagnostic Field Testing of Fluid-Filled Power Transformers, Regulators, and Reactors.
Ravi Raj Gupta is the Substation Applications Engineer at Megger Americas. His responsibility includes providing technical and engineering support for various testing equipment manufactured at Megger. His area of focus is testing equipment that primarily helps test transformers, circuit breakers, and other substation assets. He previously worked for Siemens as a Lead Application Engineer and a Business Development Manager for the voltage regulator product line within their Transformers Division. Gupta received his MS in advanced electric power engineering and electrical engineering from Michigan Technological University and holds an Engineer in Training certificate from Michigan.
Swapnil Marathe is a Substation Application Engineer with Megger. He worked as an intern with Megger on relay and substation applications from August 2019 to May 2020 and started as a full-time employee in May 2020. Marathe focuses on testing transformers, batteries, circuit breakers, and motors. He earned a BS in Electrical Engineering with a major in Power System studies from the University of Texas at Arlington and is an active IEEE member involved in the Transformer Committee and Battery Committee.