Testing of Transformers

RATIO, POLARTITY AND PHASE RELATION TESTS

The purpose of a transformer is to convert power from one system voltage to another. This voltage relationship, or voltage ratio, is determined by the number of turns on the high voltage coil and the number of turns on the low voltage coil of the transformer. The ANSI specification states that the measured ratios must have a maximum variation of plus or minus 0.5% compared to specified ratios.

The ratio is measured by using a transformer turns ratio-meter (commonly called a TTR) which applies a low voltage to the transformer under test. A comparison is made between the transformer under test and an adjustable variable ratio transformer in the ratio-meter, when the two ratios' are equal; a balance is obtained on the detector. A high exciting current during the ratio test may indicate a shorted turn.

For three phase transformers, the polarity and the phase relationship between the high voltage and the low voltage should to verify to the design requirements. These characteristics are described in diagram 1 and are particularly important when two or more transformers are paralleled. The phase relationship is the angular phasor (vector) displacement between, say, H1 and X1 and is typically 30 degrees leading.

RESISTANCE MEASUREMENT

The measurement of the dc resistance of the transformer windings and comparison to the design calculations verifies that the correct conductor size has been used, brazed, crimped or bolted connections are satisfactory and that the contact resistance of any tap switches or tap changer is within acceptable limits.

The measurements must be made at a known temperature, at least 24 hours after oil filling, and corrected to the standard reference temperature. An accurate measurement is necessary since it is used to correct the load losses to the reference temperatures as discussed under Load Losses and to calculate the winding temperatures after a temperature rise test.

NO-LOAD LOSS AND EXCITING CURRENT TESTS

The load losses and the exciting current are measured by using connections as shown in diagram 2, rated voltage is applied to either the high voltage or low voltage side while the other side remains open circuit.

By applying an alternating voltage to the one side, a magnetic flux is established in the core, which induces or causes a voltage to appear across the terminals of the other side. The exciting current and no load loss or core loss is the energy required to establish (or excite) the magnetic flux in the core. To obtain accurate values of no load losses, the wave shape of the applied voltage must be a close as possible to a sine wave. A correction for the wave shape variation is made to the measured results.

These tests are performed to ensure that the electrical performance of the core is comparable to the calculated values. They verify that the core has been designed and built correctly, that the quality of the core materials is satisfactory and the core is operating in the correct range of flux density. Transformers are designed to operate near or below the knee of the magnetic performance curve for the core steel material as illustrated in diagram 3. This practice avoids the core operating into the saturation point of the magnetic performance curve which would cause the no load loss and magnetizing currents to increase sharply. It also enables the transformer to operate in accordance certain over voltage conditions as specified in the ANSI Standards without exceeding its rated temperature rating.

Occasionally a user or specifier will request that the no load loss and magnetizing currents are measured at 90%, 100% and 110% rated voltage. These tests verify that the core will not operate into saturation during the ANSI specified over voltage conditions.

LOAD LOSS AND IMPEDANCE TEST

During the load loss and impedance test, a voltage is applied to the high voltage side of the transformer while the bushings of the low voltage side are shorted together as shown in diagram 4. An applied voltage is increased until the current supplied matches the rated current. Losses supplied to the transformer under these conditions are equivalent to the load losses that are incurred during full load operation.

Load losses are the sum of the resistive losses in the windings plus stray losses in the tank, core clamps and other metal parts and eddy losses due to circulating currents in the winding conductors. The resistive part of the load losses is sensitive to temperature since the winding resistance increases with increasing temperature. On the test report, the losses are corrected to either 75 degrees C for 55-degree rise units or 85 degrees for 65-degree rise units, which are the specified reference temperatures.

The applied voltage required to supply the rated current during this test is used in the calculation of impedance voltage. Impedance is expressed as the percentage ratio of applied voltage to rated voltage. ANSI Standards allow impedance tolerances for two winding transformers of plus or minus 71/2% and for transformers having three or more winding, auto-transformers or grounding transformers the tolerance is plus or minus 10%.

The impedance of a transformer determines the amount of fault current flowing in the windings should a short circuit occur during the operation of the transformer. The magnitude of short circuit currents flowing through the transformer, assuming no system impedance and an infinite supply, would be the rated current times the reciprocal of the per unit impedance. An impedance percentage of 5% or 0.05p.u. will cause a potential short circuit current of 20 times the rated current. The lower the impedance the higher the potential short circuit current.

The mechanical forces acting within the transformer are directly related to the square of the magnitude of current flowing through the windings. The designer calculates the design's capability to withstand these mechanical forces and, if necessary, makes adjustments to ensure the transformer meets the ANSI criteria. Sunbelt Transformer designers uses a combination of several methods based on many years of industrial test and field experience, including "Westinghouse" calculations and Andersen programs, to confirm compliance.

The measured impedance also determines suitability for paralleling with existing transformers of known impedance. A transformer whose tested impedance is higher than the other will cause the other transformer to carry more than its equal share of the load.

DIELECTRIC TESTS

Dielectric tests are the group of tests during which the transformer will be subjected to higher voltage levels and therefore higher voltage stresses than would normally be experienced in service. The purpose is to confirm that the design, manufacture and processing of the transformer and insulation structure and materials are adequate to provide many years of satisfactory life.


APPLIED POTENTIAL TEST

The applied voltage test (commonly called Hipot test) verifies that the major insulation structures and the clearance between leads and ground are satisfactory. The major insulation is the insulation between the winding under test and the other windings, to the core, to core clamps and tank. The test level for each system voltage is specified in ANSI and is applied at power frequency for one minute.

During this test, each winding is shorted out by connecting its bushing as shown in sketch 5, the specified voltage is then applied to the winding under test with the other windings connected to ground. No voltage is induced in the winding under test or magnetic flux induced in the core; hence the insulation between turns or between layers in the winding under test is not stressed.

INDUCED VOLTAGE TEST

The induced voltage test, diagram 6, induces a voltage in the transformer and causes the voltage stress between turns and between layers of each winding to be raised to higher than normal service voltage. The applied voltage is typically at 120 or 180 hertz in order not to saturate the core that would occur if it were at power frequency. To allow for variation in test frequencies, the test duration is 7200 cycles.

SOUND LEVEL TEST

The sound level is measured by supplying one side of the transformer at rated voltage with the other side open circuit, diagram 2. The sound level is measured at a specified height and positions around the transformer and the results averaged. It is necessary to provide a low level of background sound and avoid reflective surfaces to achieve accurate results.

For particular applications, users may specify lower levels than those in ANSI standards. Since the core is the source of sound, the noise can be reduced by lowering the flux density and by mounting the core on vibration pads in the tank. However, by reducing the sound level in this way, the initial cost of the transformer will be increased while the no load losses will be reduced.

For forced cooled transformers, the fans are a source of sound, low speed fans or fans with special blades can be used to reduce sound levels.

TEMPERATURE RISE TESTS

The temperature rise test is a way of verifying the cooling of the transformer and is performed by using the connections show in Sketch 4. The temperature rise test is performed in two stages, first by the supply of total losses, the oil rise temperatures are established. The end point is reached when the rate of top oil temperature rise is essentially flat. The supplied power (current) is then reduced to provide the load losses which would occur in the highest loss tap position and the test continued for one hour. At this point, the power is isolated, the shorting connection is removed and the winding resistances measured during the next 10 - 15 minutes. A curve of the reducing resistance against time is plotted, called a cooling curve, and is extrapolated back to time zero to provide the winding resistance at the instant of shutdown. Calculations are made to give the temperature of the windings during the time they were carrying rated load losses. Under certain conditions, the standards allow for one reading of winding resistance to be taken and a correction made.

The temperature rise tests can typically take around 10 to 20 hours and are considered an optional test. Typically a user will accept the results of a similar unit instead of performing a test.

The life expectancy and rate of aging of paper and pressboard insulation used to insulate the windings depends on its service temperature, time at that temperature and the condition of the surrounding oil. In order to achieve normal life expectancy of the insulation materials and hence the transformer, the ANSI and NEMA specify temperatures that the transformer must not exceed when delivering its rated output.

The ANSI standards provide information on the loss of life of the insulation should it be operated above the specified temperatures. For every 8 degrees C above the specified insulation temperature limits, the rate of losses of life is doubled. For users that require to overload their transformers from time to time, this guide provided useful operating information.

CORONA (PARTIAL DISCHARGE)

Corona or partial discharge is measured during the induced test using special equipment coupled to the transformer bushings. The level of discharge, measured in micro-volts or picocoulombs, is continuously monitored and recorded every 5 minutes. Transformer, rated 115KV and above, require this one hour test during which all parts of the insulation is overstressed to 150% of normal levels. The ANSI specifications mentioned above describe these tests and specify acceptable levels for various system voltage levels. Sunbelt Transformer is equipped to measure both micro-volts and pico-coulombs simultaneously.

During the operation of a transformer, electrical discharges may be generated which cause loss of life of the insulation materials and interference of electrical communications in the area around the location of the transformer. These discharges can be caused by several factors for example a) inadequate processing/vacuum filling that leaves air voids in the insulation or oil, b) concentrations of high electrical stress at sharp points/edges on conductors or c) local points of overstressed insulation. Experienced manufacturers have developed various techniques to minimize the likelihood of corona and pay particular attention during manufacture and oil filling processes.

INSULATION-RESISTANCE TEST

For two winding transformers, the bushings of each winding are shorted together and the tank and the core grounded. Measurements are taken between the high side to low side grounded and then between the low side to high side grounded. The core is also measured to ground if available via an external bushing or an accessible internal connection.

 The dryness, cleanliness and the temperature of the transformer will effect the value of insulation resistance. By measuring the insulation resistance, correcting to 20 degrees C, and comparing to available published data, the quality and the reliability of the transformer can be estimated. During manufacture this test is performed a number of times not only to determine dryness but also to identify unintentional shorts or ground circuits, particularly between the core steel and the core clamping structure. During routine preventative maintenance, these measurements can be taken and compared to the original values to determine the condition of the insulation.

INSULATION POWER FACTOR

The insulation power factor test is another test that can be performed to determine the condition of the transformer insulation. The measurement is made with a capacitance bridge, measuring the capacitance between windings and between windings and ground, together with the power factor or loss angle of this capacitance. The dissipation factor is a similar test that provides useful information as to the dryness and condition of the insulation. Typically, at 20 degrees C, the power factor of a new transformer should be below 0.5% and values of half this value can be achieved in a new transformer.

Again, the values obtained during the factory tests on a new transformer can be compared to those taken during routine maintenance tests and some indication of the deterioration of the insulation structure can be determined.

OIL TESTS

Several tests can be performed on transformer oil to determine its condition and many are particularly useful in determining the condition oil that has been in service for a number of years. The most common test performed on oil is the dielectric test to either ASTM D877 or ASTM D1816. These specifications define the method of preparing a sample, the equipment to be used and the test method.

During factory tests it is useful to take oil samples before testing and again after testing in order to perform a dissolved gas in oil analysis. This is particularly useful for high voltage transformers and in cases when a temperature rise test be performed.

Oil tests are a useful indicator of the condition of the insulation system and the oil and form Important elements of any transformer preventative maintenance program. Results taken after periods of service can be compared to baseline measurements taken on the new transformer.

REGULATION AND EFFICENCY

As a transformer supplies an increasing load, the actual voltage at the secondary terminals reduces and falls below the specified or designed voltage. This voltage drop or reduction is called the regulation of the transformer and is related to the impedance of the transformer, lower values of impedance will provide a reduction in regulation for a given load. The regulation is the ratio of output voltage at a specified load, compared to the output voltage at no load. Calculation of the regulation at various loads and power factors requires calculation of the two components of the impedance, i.e. the reactance and resistance established from results of earlier tests. Regulation increases as the power factor reduces or deteriorates.

Since utilities are required to maintain the system voltage within a certain range, regulation is an important operational consideration and it can be compensated for by the use of voltage taps in one or both of the winding. Typical no-load tap changer adjustment are +5% above rated voltage to -5% below rated voltage in 2 � % steps which is achieved by adding or removing active turns in the winding. For certain applications, transformers are provided with On-Load tap changer having an increased range of adjustment with the ability of making tap changes without disconnecting the load. For transformers having only no-load taps, excessive regulation can be compensated for by the installation of separate step voltage regulators.

The efficiency of a transformer is usually well over 90%. It is calculated by the ratio of the output divided by the input. Expressed as a per unit value, the calculation would be 1 - total losses divided by the input. The total losses are those that would occur at the rated input considered. Charts are available to facilitate the determination of the efficiency at various loads