Authors: Tony Wills, Team Leader Applications Engineer, and Jill Duplessis
The vital role of transformers in the power system is well known: they transfer electrical energy from one electrical circuit to another and typically, during this process, change the voltage of the electricity flowing in the circuit in a predetermined way. However, the voltage transformation that is expected to occur, and which is indicated by a transformer’s nameplate, applies specifically to the case when a transformer is energised without load. When a transformer is carrying load, the transformer’s secondary voltage will diminish by an amount determined by the transformer’s impedance and the power factor of the load. But as the served load and its power factor varies, it is important that the output voltage of a transformer stay within narrow limits as most loads require constant voltage. This desired ‘voltage regulation’ is made possible through the use of a tap changer.
A transformer tap changer facilitates control over the ratio of a transformer’s output voltage to its input voltage. It is a mechanical device and, in many cases, a transformer’s only component with moving parts. Pumps and fans are the others, but they are only present on transformers with an oil- and/or air-forced cooling system. Of critical distinction here, however, is that a tap changer is the only component with moving parts that is connected to a transformer’s winding(s)! Reliability of the tap changer is crucial, because taking a transformer out of service to deal with a tap changer problem is expensive and disruptive. Moreover, a failure of a tap changer may be catastrophic. It is, therefore, in every utility’s interest to carry out regular condition assessments on tap changers to detect developing faults before they lead to failures. After briefly discussing types of tap changers, this article looks at both static and dynamic resistance test methods that can be used to perform these essential assessments.
A tap changer changes the ratio of transformation by adding or subtracting relatively small sections of tap (or regulating) windings to or from the HV or LV main windings of the transformer. Ratio changes involve the mechanical movement of a contact from one position to another and, in assessing the condition of a tap changer, it is the performance of this moving contact; its respective stationary mating contact; additional working parts of the tap changer mechanism; and sections of tap windings that need to be checked. There are many problems that may occur in a tap changer. For example, misalignment during manufacturing and/or during transportation, looseness of the moveable contact, and contact wear all lead to insufficient surface contact, with the consequence that the full-load current overheats the contact and produces coke, a solid residue created when oil undergoes severe oxidative and thermal breakdown. Other common issues are: failure of the make-beforebreak sequence during tap changing, which leads to arcing; improper wiring; problems with the diverter/arcing switch or with transition resistors; and open-circuited or short-circuited turns in the tap windings or, in the case of reactance type OLTC’s, the preventative autotransformer (PA), series autotransformer or series transformer.
Types of tap changers
Tap changers are divided into two main types: on-load tap changers (OLTCs or simply LTCs), which allow the transformer ratio to be changed while the transformer is in service and passing current, and off-load (or deenergised) tap changers (DETCs), which require the transformer to be taken out of service before the ratio can be changed. A DETC may also be used to change configuration of the windings. By varying the transformer ratio under load without interruption, OLTCs enable voltage regulation and/or phase shifting. The job of an OLTC is impressive when you consider that an OLTC is typically required to operate and transfer load current several thousand times a year.
OLTCs can be further subdivided into resistor and reactance types. Most experience with dynamic measurements is with resistor type OLTCs, which is reflected by the examples given in this article, but its STATIC AND DYNAMIC RESISTANCE TESTING OF TRANSFORMER TAP CHANGERS ET 103 use has expanded in the past several years to include reactance types as well.
Figure 1: Resistor type OLTC
Figure 1 shows a typical resistor type OLTC with a tap selector and a diverter switch. The transition resistors are typically just a few ohms. The total operation time for a resistor type OLTC, from receiving a signal for tap switching to the tap changer reaching its final position, is usually between 3 and 10 seconds, depending on the design. The actual contact switching time for this type of tap changer is usually around 40 to 60 ms, and the transition resistors are loaded for about half this time or a little longer. Resistor type OLTCs, which are connected to the HV winding of the transformer, are most commonly used in Europe, Asia, Africa and South America.
Reactance OLTCs use a preventive autotransformer (PA) in place of the resistors used in a resistor type OLTC, which means that the additional resistance in the diverter device is very low. Switching time in reactance OLTCs is significantly longer, in the order of seconds. Reactance OLTCs are almost always connected to the LV side of the transformer and are very common in the US as well as in Canada and Mexico.
Unlike a resistor type OLTC, a reactance OLTC uses the bridging position (i.e. positions whereby two consecutive taps are selected at the same time and some form of impedance, resistive or reactive, is present to limit resulting circulating current) as a service position. Therefore, when performing a static measurement on an odd numbered/bridging tap position of a reactance OLTC, the bridging components (e.g. the PA) are included in the test circuit and can be assessed. Since a resistor type OLTC never ‘rests’ in a bridging state, when performing a static measurement, its bridging components, such as diverter resistors, are not included in the test circuit no matter at which tap position a test is performed. Therefore, dynamic measurements are critical to assess these ‘transition components’, in addition to providing other diagnostic benefits.
Tap changer test methods
Winding resistance measurements (static)
There are a number of static measurements that can be performed while the OLTC (and DETC) is stationary at each of a number of selected tap positions, including excitation current, transformer turns ratio, winding resistance, and SFRA tests. DC Winding Resistance Measurements (WRMs) are normally performed for every tap in the same way that they are carried out for individual windings. The test instrument injects current continuously and the resistance for each tap is measured sequentially as the tap changer is stepped through its positions. The results are typically presented as a graph or table. Resistance changes between taps should be consistent, with only small differences between phases and tap positions.
There are several techniques for carrying out dynamic measurements on tap changers but common to all of them is that DC current is injected into the tap changer, either in one phase or in all phases, and the current and/ or voltage is measured as a function of time during the operation of the tap changer. Test currents range from about 0.1 A DC to the standard test current for winding resistance measurements, which is typically from 1 % to no more than 15 % of the rated current for the transformer winding. Dynamic tests can be performed at the same time as WRMs, or as separate tests. Standard tests are:
Figure 2: Dynamic current measurement
- Continuity/make-before-break testing
- Dynamic measurements
- Dynamic current/ripple (often called ‘DRM’ even though focus is completely on the behaviour of the test current)
- Dynamic voltage (often, the end goal here is to produce a DRM plot) n
- Dynamic resistance (DRM) – a new approach to most accurately capture the actual DRM of each tap change operation
- Motor current
Continuity tests/discontinuity detection
By monitoring the current or voltage change, this test detects if there is a break-before-make condition in the tap changer. If there is, the current will be interrupted during the operation of the tap changer. This test is typically carried out at the same time as winding resistance measurements.
Open contact detection can be performed by looking for current changes, changes in di/dt, or by detecting voltage changes on the generator output or opposite side of the transformer. In other words, if the tap changer is on the HV side, detection is performed by measuring voltage transients on the corresponding LV winding.
Dynamic current measurement
Dynamic Current Measurement (DCM) is in some ways similar to continuity testing but, in addition to simply detecting discontinuity, the current is monitored throughout the tap change and the result is presented as a percentage ripple value or as a current-time graph (Figure 2). Ripple is the magnitude by which the test current decreases during the tap change and is expressed as a percentage of the test current. The slope is also evaluated when a current-time graph is provided. This reflects the speed at which the test current decreases when the moving contact breaks from a stationary contact and the current flows entirely through one transition resistor. The aim of the DCM test is to look at the conditions during the tap changing operation and provide information about contact timing.
Figure 3: Test set-up for performing DRM on an OLTC: current is injected in HV winding; voltage is measured on LV winding
Dynamic current measurements are affected by the magnitude of the test current. If the test current is below the saturation level of the transformer, the inductance of the transformer winding is high, and this smooths out current changes. If the test is performed with a current at or above the saturation level, the inductance will be low, and the current changes will be greater. A way of reducing transformer inductance when performing measurements on tap changers is to short circuit the ‘not directly under test’ LV (or HV) windings. This, in effect, replaces the inductance of the winding with the shortcircuit impedance. Inductance is greatly reduced and changes in current can be measured more accurately. The disadvantage with this shorting method is that DCM cannot be performed together with winding resistance measurements because static winding resistance tests must be performed with the opposite windings opencircuited.
OLTC dynamic current measurements are typically performed during winding resistance test ‘rest times’, that is, as the tap changer moves to each successive tap position, and indications of discontinuity are monitored.
Test currents for HV winding resistance measurements are often around saturation level or higher, so the inductance of the winding is already low, as desired. Note that this method of testing may show large current changes/ ripple even for good contacts. Ripple should be similar, however, for the same tap change in each of the three phases.
Dynamic voltage measurement
Dynamic Voltage Measurement (DVM) is an early method to determine DRM. A relatively small and constant current from a high impedance source is injected through the tap changer and the voltage over the test circuit is measured with sufficient resolution to create a resistanceversus- time diagram for the operation of the tap changer contacts. Interruptions are easily recognised, changes in the diverter resistor values can be assessed and contact timing can be measured.
An example of the test set up for performing DVM on an OLTC is shown in Figure 3, and the results for an OLTC in good condition are shown in Figure 4.
‘True’ dynamic resistance measurements
A new, patent-pending method is to measure dynamic resistance in the on-load tap changer by simultaneously measuring the test current together with voltages on both HV and LV windings and combine the results with transformer modelling. An example of a measurement is shown in Figure 5.
Figure 4: DVM test results on a good OLTC
The source impedance in this example is about 10 Ω and we can see a small current change during the tap change (green trace in Figure 5). Due to the inductance in the circuit (recalling that, for this measurement, the opposite or LV winding is left open), the change in voltage measured across the HV winding (red trace in Figure 5) is rather large. This voltage is a sum of inductive and resistive voltage and cannot be used for directly calculating the resistance in the circuit. However, the voltage measured across the LV winding is purely inductive and if we use transformer model parameters to calculate the inductive voltage on the primary, we can deduct this value from the measured HV winding voltage and calculate the resistance in the circuit. The result is given in Figure 5.
Figure 5: DRM on OLTC using a 5 A test current and 10 Ω source impedance [test current (green); voltage measured across the HV winding (red); voltage measured across the LV winding (blue), resistance (black)]
Figure 6: Switching sequence for a typical OLTC with diverter switch, Maschinenfabrik Reinhausen GmbH, 2002
Figure 7: DRM test on a MR type V OLTC, test current: 5 A, normal condition
The advantages of this true DRM method over the DCM test are that inconsistencies in test results, for example those that are often introduced by selection of current source, are removed and transition resistor values can be determined.
Motor current measurements
The method used to measure the tap changer motor current is basically the same as the method used to measure the coil current in a circuit breaker. The motor current during a tap change reflects the energy used by the operation, and the measurement can be used as a fingerprint for trending and benchmarking.
Analysing the results
Winding resistance measurements
Measuring the winding resistance for each individual tap is quite straightforward. The most commonly encountered problem is that the person carrying out the test has not allowed sufficient time after a tap change before taking a measurement. The resistance value should be observed closely to make sure it has stabilised before the value is recorded and stored. At each tap position, resistance values for each phase should be compared. If all readings are within 1 % of each other, then they are acceptable. Deviation should not exceed 3 %.
There should be no discontinuity indication for any tap change. If a discontinuity occurs during the test, there will be a large increase in the measured voltage and a large decrease in the measured current.
Switching times are critical for the correct operation of tap changers. Figure 6 shows the switching sequence for a typical design. As can be seen, the total switching time is around 60 ms for a resistor type OLTC and is made up of a number of steps (plus the static positions before and after the tap change operation). Figure 7 shows the results of a DRM test on a similar type of OLTC.
Note that this type of tap changer has six distinct states for each tap switching operation, all of which are recognisable in the DRM test results.
These states are:
- Load contact releases
- R1 inserted
- R2 inserted in parallel with R1
- R1 released
- R2 released
- Load contact makes
Contact bounce can be seen especially in relation to R2 making. The total time is 79 ms, and the resistor switching time is 47 ms. For comparison, Figure 8 shows results for the same tap changer with a spring removed to simulate spring damage. Measured switching times are about twice as long.
Figure 8: DRM test on a MR type V OLTC, test current: 5 A, simulated spring fault
Figure 9: Motor current envelope; normal (blue), simulated high friction (red)
Switching times for an OLTC in good condition should be consistent over all taps. Dynamic voltage measurements on the LV side of the transformer (assuming a resistive type OLTC on the HV side) can also be used to perform timing measurements, an approach which has the benefit of simplicity. Dynamic current measurements are another viable option for measuring tap changer timing, but the steps look different from each other and are sometimes not as distinct as when dynamic voltage or resistance is used.
Transition (or diverter) resistor values
The values of the diverter resistors should be within 10 % of their nominal values.
The motor current should be compared with previous test results or with the motor current measured in similar units. Increased friction in the OLTC mechanism leads to higher motor current because the motor is required to develop more torque to operate the mechanism. This can be detected by making measurements of the motor current over extended time periods and comparing each measurement with the expected normal shape. Figure 9 shows an example of the motor current curve where a simulated increase in friction causes an unusually high motor current around the time the springs are charged before the main switching operation.
As one of the few moving parts in power transformers, OLTCs are the parts most prone to wear and ultimately failure. The results of such failures can be very costly, but much can be done to guard against them by regularly evaluating OLTC performance using the test and measurement techniques outlined in this article. Megger offers instruments that can be used to carry out these tests and will be pleased to advise on selecting the most suitable of these for specific applications.