Transformer & Substation Testing with Portable Equipment
- March 20, 2020
- Posted by: Dominique
- Categories: Substation Testing, Transformer Testing
From the first day of use, transformers are subject to thermal and mechanical stresses, foreign particle ingress, and variations in temperature and humidity. HV TECHNOLOGIES, Inc. offers the top tier range of portable instruments from Haefely / Tettex for testing transformers onsite, in the factory, and in the repair shop.
Power Factor / Dissipation Factor & Capacitance:
The power factor/dissipation factor (cosφ/tanδ) and capacitance measurements help to determine the insulation condition in bushings or between windings. Power factor/dissipation factor is the insulation loss factor. “Power Factor” cosφ is used in the United States to describe dielectric losses, while “Dissipation Factor” tanδ is the calculation used in Europe.
At normal operating voltage and frequency there is little loss in all insulators used in a transformer. This loss changes in direct proportion with the “square” of the applied voltage. To specify the insulation loss factor, the test object must be considered in the test arrangement as a capacitor. Considering a simple disc capacitor, the capacitance C can be calculated by C = (A x ε)/d where A is the surface area of the electrode face, d is distance between the electrodes, and ε is the dielectric constant of the material.
For simplistic purposes, an insulating material can be modeled by an equivalent circuit consisting of 2 elements; a resistor and a capacitor. When voltage (U) is applied to the device under test the total current (I) will be the contributions of the resistive current (IR) and capacitive current (IC). Tan delta is the ratio between the resistive current and capacitive current. If we consider the resistance of an ideal material to be infinite, then the total current will be purely capacitive and 90° phase shifted to the applied voltage. However, in real systems, losses will allow for resistive currents to flow, which will have an impact on the phase shift of the total current with respect to voltage. As a result, a loss angle (δ) will appear, and the tangent of this will correspond to tan delta, and the cosine of this will correspond to power factor, both of which are dimensionless units.
The loss angle δ depends heavily on the thickness of the insulation material, surface condition, structural property of the insulator, type of material, and also on the humidity, number of foreign materials/particles, air gaps, etc. which causes ionization within the insulating material. The conditions that increase power losses of insulating material also decrease the insulation strength. Therefore, loss angle measurement provides valuable criteria for evaluating the insulation material at a defined operating frequency.
Winding resistance can be performed as a type test, routine test, and a field test. It is used to check transformer windings and terminal connections and is also used as both a reference for future measurements and to calculate the load loss values at reference temperature. Measuring the winding resistance is done by using a DC current and is very much dependent on temperature. The temperature is corrected using the equations below:
Due to this, the temperatures must be measured and recorded when measuring the winding resistance. The Haefely 2293 Winding Analyzer has 6 temperature channels with automatic resistance correction. These can be extended up to 30 by using the optional temperature extension box (2293/TEMP), which can be used with available magnetic and liquid temperature probes.
The 2293 consists of three programmable power supplies, which can operate in constant current or constant voltage mode. Further, these power supplies operate in two quadrants. This means that a power supply can act as an active load which can be used to quickly discharge the current from high inductive DUTs. To determine the resistance of the DUT the 2293 uses two voltage and two current measuring channels. This way the unit can measure two different resistances with individual currents at the same time. The resistance calculation is according to the principle described in the figure below, and the reading is conducted once the current has stabilized.
Equipment for winding resistance testing: 2293 Winding Analyzer
“Turns Ratio” (TR) is defined as the theoretical voltage ratio, the term “Voltage Ratio” (VR) as the ratio of the rated voltages (“boilerplate voltages”).
The goal of the measurement is to confirm the no-load voltage ratio given in the specifications, determine the conditions of both the windings and the connections, and searching for problems (if any). Turns ratio can either be measured by the bridge method or by measuring the voltage ratio of the windings. The Haefely 2293 Winding Analyzer, TTR 2795 and TTR 2796 utilize the latter.
As a rule, the voltage ratio of three-phase transformers can be measured using a single-phase supply as long as the distribution of magnetic flux in the core is taken into consideration. Only windings, winding segments, and winding combinations which have the same magnetic flux applied can be compared with one another. The measuring circuit can be derived from the phasor diagram of the test transformers’ vector group. The two voltages being compared must be in phase and have the same orientation.
However, depending on the winding configuration and phase rotation of the transformers, it is not always possible for these voltages to be in phase. Therefore the 2293 and the TTR 2795/2796 internally correct the measurement with the “TR/(VA/Va)” factor and automatically display the correct ratio value. The TR/(VA/Va) factor determines the turns ratio from the ratio of the energizing voltage to the output voltage of the winding. The ratio of the measured voltages is multiplied by this factor to give the turns ratio for that winding/tap.
Due to similar reasons the voltage ratio VR and the turns ratio TR are (depending on the winding configuration) not always equal. Therefore the 2293 and TTR 2795/2796 internally correct the entered voltages (VR) with the “VR/TR” factor and automatically display the correct ratio deviation which is calculated from the rated voltages (boilerplate voltages) as reference and the measured (and corrected) ratio value. The VR/TR factor determines the multiplier applied to the measured voltage ratio (VR) to get the turns ratio (TR) of the transformer. When the boilerplate voltages for the transformer are set up the calculated voltage ratio is divided by this factor to give the nominal turns ratio for the winding/tap.
The main reasons why insulation fails are excessive heat, moisture, dirt, vibrations, and aging. Insulation resistance measurements are made to determine the insulation conditions of the transformer’s windings to earth, between windings and to form a reference for future measurements during operation.
The currents flowing in the resistance formed by the insulator consist of the leakage, charging, and absorption currents. They are heavily dependent on humidity of the insulator, foreign materials in the insulator, and temperature. These currents are also time dependent. At the start of the measurement the current is dominated by the capacitance charging current. Roughly one minute later the capacitance is charged and the total current consists primarily of the absorption current, which is caused by molecules that polarize. The absorption current decreases to a negligible value after a few minutes, and if no surface leakage is present, the current flowing now consists only of leakage current, which should be stable.
Insulation resistance of a transformer should be done using DC voltage and should be between 1000 V DC and 5000 V DC.
After disconnecting a transformer from the grid or performing a winding resistance measurement with direct current, the transformer core will be magnetized. The figure below shows a transformer core hysteresis curve with a possible magnetization M0, which can be anywhere on the y-axis within the hysteresis loop.
The magnetization M0 can influence various measurements like turns ratio or frequency response. For these measurements the magnetization should be M0≈0Am-1, otherwise the results can be incorrect or not comparable. Further, connecting a magnetized transformer to the grid can cause high inrush currents.
The common method to demagnetize a transformer core is to apply nominal AC voltage to the transformer and slowly decrease its amplitude to zero. This method requires a very large and non-portable controllable AC voltage source. The demagnetization function within the 2293 Winding Analyzer calculates the hysteresis loop of the transformer using a special algorithm and then an iterative procedure is run in several cycles until the demagnetization status is reached.
Equipment for demagnetization testing: 2293 Winding Analyzer
The short-circuit impedance is defined as the impedance corresponding to the voltage that must be connected to one pair of terminals of a transformer, with terminals of the other side shorted, causing the rated current to flow on both the HV and LV side of the transformer. The equation for this is ZCC = UCC / Irated
For single phase transformers, the short circuit impedance in % can be easily calculated after the ZCC has been measured according to the formula εCC = ZCC * (S/U2) * 100, where S is the apparent power in VA from the transformer nameplate.
The Haefely 2293 Winding Analyzer and MIDAS 2881G can both perform the short circuit impedance test. The 2293 requires no additional hardware, while the MIDAS requires the use of the “Current Booster 5287” to provide an output current up to 10 A (15 A intermitted) at low voltage (10 V or 100 V).
Arbitrary Phase Shift:
Arbitrary phase displacement (those that don’t follow the 30° clock steps) between the primary and secondary winding are a usual feature in multi-winding transformers (e.g rectifier transformers) to reduce harmonics injected into the system (i.e. increasing the number of rectifier phases to increase the number of pulses in the rectified DC signal). The Haefely 2293 Winding Analyzer and TTR 2796 achieve arbitrary phase shift measurement by using a single-phase supply with no extra hardware. The necessary three phase equivalent voltages at different transformer terminals are simulated by merging a pair of transformer terminals and interconnecting certain other terminals. The resulting voltages at different terminals are used to determine the arbitrary phase shift.
The excitation current of a transformer is the current required to energize the core. Even with a zero load, a transformer will draw a small amount of current due to internal loss. The excitation current is made of two components – a real component in the form of losses that are commonly referred to as no-load losses and reactive power measured in kVar. The excitation current measurement can be used to detect short-circuited turns, poor electrical connections, core delaminations, core lamination shorts, tap changer problems, and other possible windings and core problems in the transformer. In principle the test measures the current needed to magnetize the core and generate the magnetic field in the windings.
The excitation current test should be performed before any DC tests. Excitation current tests should never be conducted after a DC test has been performed on the transformer, as results will be incorrect because of residual magnetism of the core left from the DC tests.
Heat Run Test (Temperature Rise & Cooling Curve):
The purpose of this test is to establish the top oil temperature rise in steady-state condition with dissipation of total losses and to establish the average winding temperature rise at rated current and with the top oil temperature rise specified before. This target is achieved in two main steps. First the transformer is subjected to a test voltage such that the active power is equal to the total losses of the transformer. The top oil and average oil temperature rises are established in this phase. The oil temperature and the cooling medium temperature are monitored, and the test is continued until a steady-state oil temperature is reached. When the top oil temperature has been established, the test shall immediately continue with the test current reduced to the rated current for the connected winding. This condition is maintained for a certain time, typically one hour.
Following the heating step, the resistance has to be measured after a quick disconnection of the power supply. The values of average temperature of the two windings are determined from the resistance change over time, known as cooling curve.
Equipment for heat run testing: 2293 Winding Analyzer
Tap Changer Dynamic Resistance:
The tap changer dynamic resistance test records currents and voltages during the tap position transition. A tap changer is a connection point selection mechanism along a power transformer winding that allows a variable number of turns to be selected in discrete steps. A transformer with a variable turns ratio is produced, enabling stepped voltage regulation of the output. The tap selection may be made via an automatic or manual tap changer mechanism.
If tap switching can be done while current is circulating (transformer connected to the grid) the tap changer is called an On-Load Tap Changer (OLTC). Simple changing of taps (switching) during an energized condition is unacceptable due to momentary loss of power during the switching operation. Therefore the “make before break contact concept”, is the basic design for all OLTCs. This concept, which means the next tap is selected before disconnecting the previous one, would generate a large short circuit current between turns due to the large number of Ampere-Turns in a few turns. To limit this short circuit current, an impedance (resistive or inductive) must be inserted between the two taps to be switched.
A standard current curve while measuring dynamic resistance would show the following shape:
The parameters in the below graph are calculated and recorded by the Haefely 2293 Winding Analyzer for each tap change. tfall (in ms) is the time from the starting of the tap changer movement (drop in current) until the minimum current is reached. trise (in ms) is the time from minimum current reached until the current becomes stable again. The delta is the current drop from stabilization of current after it has dropped to a minimum during a tap change.
Equipment for tap changer dynamic resistance testing: 2293 Winding Analyzer
The magnetic balance test is conducted on transformers to identify inter-turn faults, magnetic imbalance or the flux distribution. This is particularly useful for checking the balance after a DC test current has been applied to the transformer during a resistance measurement which may have caused the core to become magnetized. It can be conducted on either the HV or LV side of the transformer. On the HV side, a voltage is applied across 2 phases (say 1U and 1V) and the voltage across the other phases is measured (between U-W and V-W). The test is repeated for each of the three phases. The result is the sum of the two measured voltages will be equal to the applied voltage, such as 1U1V=1U1W+1V1W. The voltages obtained in the secondary will also be proportional to the voltages in the primary, indicating that the transformer is magnetically balanced. This helps to identify if there is any inter-turn short circuit that may result in the sum of the two voltages not being equal to the applied voltage.
Equipment for magnetic balance testing: 2293 Winding Analyzer
Sweep Frequency Response Analysis:
The evaluation of the frequency response (transfer function) can be used on-site to detect defects, in particular displacements in transformer windings. These types of defects can occur during transportation or during a short circuit near the transformer in the power system. These defects can also occur after a long service life of transformers in power systems, where several short circuits and over voltages over the life time may happen.
The corresponding HF circuit of a transformer winding is a complex R-L-C network. The measured frequency response of this network is unique like a fingerprint. The frequency response describes the frequency behavior of a two-port network. To determine the frequency response of a transformer, the assembly is considered as a linear, complex, time-resistant and passive network. One input signal UIN and one output signal UOUT or IOUT have to be defined. The division of the two signals represent the frequency response. The following three transfer functions are measured and calculated:
Transfer Voltage Function UOUT / UIN (f)
Transfer Impedance Function UIN / IOUT (f)
Transfer Admittance Function IOUT / UIN (f)
The frequency response method is normally a relative method of diagnosis. Actual measurements have to be compared to reference results (previous measurements). Results of frequency response analysis (FRA) measurements are conveniently displayed in modulus-argument form. The modulus is usually referred to as the gain or the amplitude and the argument as the phase.
There are 3 types of comparisons when speaking of FRA measurements. These are:
- Time-based Comparison: Actual results of measurements are compared to fingerprints of former tests. Unfortunately, fingerprint measurements are rarely available. This method is also used for fast detection of transportation damages.
- Type-based Comparison: The results of an identically constructed transformer (aka “Sister Unit”) are taken as a reference. Distribution transformers are especially constructed according to standards, so the results of different distribution transformers can be compared.
- Construction-based Comparison: Comparing the results of separately tested phases or coils of a transformer. Sometimes, the frequency behavior of the windings U, V and W of a 3-phase transformer are quite similar, so the results can be compared.
With type-based and Construction-based comparisons, design-based differences in the compared curves will be seen and are sometimes hard to classify. A previous measurement (fingerprint) of the exact same winding configuration (time-based comparison) is always preferred. There is no such thing as a “typical” measurement, so measurements made on “similar” transformers cannot be used as a basis for detailed comparison.
IEC 60076 standardized the measurement of frequency response in power transformers. This standardization called for data to be recorded in the XML format. The Haefely FRA 5311 adheres to this format and includes analysis software which can analyze measurements taken by other manufacturer’s devices as well, even if they’re not in the XML format. Below is a screenshot of the Haefely FRA 5311 analysis software. This software can be used even when a measurement device is not connected. Up to 10 curves can be plotted at once for convenient comparison.
Equipment for sweep frequency response analysis testing: FRA 5311
Recovery Voltage Method:
The recovery voltage method is based on the phenomenon of the polarization of oil/paper impregnated insulation. Polarization of the insulation can cause a painful shock to workers. High voltage capacitors (which were previously charged with direct voltage and even measure 0 V) can still cause a shock when the connectors are touched while being short-circuited.
There are different types of polarization. In the case of moist oil-paper insulation, there is a polarization due to the water molecules contained in the insulation. By applying a DC voltage, these molecules, which were previously electrically neutral, acquire a polarity and try to drift in the direction of the electrical field. After short circuiting the insulation some energy will still be stored in the molecules, creating a voltage that can be measured, known as the “recovery voltage”.
Equipment for recovery voltage testing: RVM 5462b