AMforum Journal 4
A magnetic core of power transformers is found very often in a magnetized state. This magnetism is called remanent magnetism, and it has undesirable effects on transformer operation. Several reasons cause remanent magnetism:
From Lachman's paper  - Influence of Magnetic Viscosity on FRA graphs
Kovan et al. in their paper  conclude that when the mitigation technique was used, the inrush currents calculated at the head of the feeder were reduced by about 60%. The IEEE PC57.102 recommends removing remanency before putting a power transformer back in service.
Several methods of removing remanent magnetism are described in the literature. The oldest ones back in 1960ties and 1970ties were performed using car battery, creating a large arc as the cables were disconnected from the transformer bushings. This method would demagnetize one phase and magnetize the other phase of a three phase unit. Also, the arc would be a scary event, that could cause an accident if not from the arc itself , then by the fall from a transformer top. This method, most commonly used for demagnetization of a transformer is based on the standard found in IEEE 62-1995 (section 188.8.131.52) (6) which directs one to alternate the polarity of a fixed voltage with decreasing application time per alternation of polarity. With each alternation, the voltage is applied until the current flow has reversed and is “slightly lower” in absolute magnitude than the current in the previous application. The time was used as a measure of the current magnitude, given the car battery voltage is constant.
Makowski in his thesis  lists three methods: Permeability Method, Time Based Method, and Integration Method. We will list other methods as well:
A. Permeability Method was expected to be the
most direct and quickest demagnetization method since it only required the
voltage to be applied once for saturation and then reversed once for
demagnetization. However, its accuracy and effectiveness regarding
demagnetization is dependent on many assumptions about the properties
of the transformer.
B. Time based, where the magnetic flux is
directly proportional to the amount of time that a constant voltage is applied
to the winding, by measuring the time needed
for the magnetic state of the core to switch from being saturated in one direction to becoming saturated in the opposite direction we can determine how long a constant
voltage needs to be applied to the winding that is saturated in order to reach the neutral point.
C. Integration method calculates the integral of the voltage applied in time, and provides as a result the exact time required to demagnetize the core applying reversed polarity.
D. The Controlled Current method is the "engineering" approach to the IEEE standard directive to alternate polarity and lower the flux in the magnetic core by controlling the current amplitude in each step. The value of the initial amplitude is not important as long the saturation is reached on all three legs of the magnetic core. Following initial current charging and discharging, the next alteration is of 60% magnitude and when reached, discharge is initiated and the steps go on in succession until the very small value of Ampere-turns is applied as the very last step.
Demagnetization process explanation (left) and test results recorded (right)
E. In the paper by Kovan  a different
procedure from the Integration one explained under C above is given. The steps
to demagnetize a core are as follows:
1) Bring the core to saturation by applying a positive voltage. This step is needed because the initial residual flux is not known. The indication that the core has
reached (positive) saturation is when the current stops increasing.
2) Reverse the applied voltage (to negative ) and measure the time that it takes for the core to be fully saturated in the reverse direction. The time to bring the core from positive saturation to negative saturation is "T". According to Faraday’s Law, the integral of the voltage from [+] to [-] gives a flux equal to twice the saturation flux.
3) Reverse the voltage once again to apply positive Vdc . Theoretically, if we apply the voltage for a time equal to "T"/2 starting from negative saturation , the core would be completely demagnetized.
Demagnetization of 1100MVA transformer following the Controlled current method
EVALUATING DEMAGNETIZATION SUCCESS
To verify that the demagnetization was performed successfully, one has two options: to perform the FRA test and compare the graph with the one obtained in the demagnetized state, or to perform the single phase Excitation current measurement (Iex) and compare the values of test current phase per phase to the known value. In case no known values are available, looking at the Iex pattern and knowing the construction of the magnetic core may be helpful.
The single phase excitation current values for a demagnetized three phase transformer with three legged core form magnetic circuit should compare between outer phases (A and C). When magnetized, these currents measured at 10kV test voltage, can show over 25% difference .
Two graphs are shown in figures below obtained with FRA method, the one where the first peaks below 1 kHz are shifted due to the remanent magnetism, and after the successful demagnetization the peaks of the outer phases do coincide as expected.
FRA traces of a 80 MVA transformer in magnetized state
FRA traces after a successful transformer core demagnetization
There are effective ways to remove remanent magnetism, even for the largest power transformers . This will extend transformer life by simply avoiding large mechanical stress at a startup, and will provide a more reliable test results when performing condition assessment using various A.C. test methods.
 Frequency Response Analysis of Transformers and Influence of Magnetic Viscosity, M.Lachman et al., Doble Conference 2010
 Mitigation of Inrush Currents in Network
Transformers by Reducing the Residual Flux With an Ultra-Low-Frequency Power
Baris Kovan, Francisco de León, Senior Member, IEEE, Dariusz Czarkowski, Member, IEEE, Zivan Zabar, Senior Member, IEEE, and Leo Birenbaum, Senior Member, IEEE; IEEE TRANSACTIONS ON POWER DELIVERY Paper no. TPWRD-00317-2010. Digital Object Identifier 10.1109/TPWRD.2010.2102778
 Proposal and Analysis of Demagnetization Methods of High Voltage Power System Transformers and Design of an Instrument to Automate the Demagnetization Process, N.J. Makowski, Masters Thesis at Portland State University 2011
 Condition Monitoring Unit (CMU), Asset maintenance department report, TNB Malaysia, 2011
 H. Kristensen,V. Mrdic, “Comparative analysis of three phase and single phase dynamic resistance measurement results,” C I R E D 22nd International Conference on Electricity Distribution Stockholm, June 2013, Paper 0473
 Report of the CIGRE Working Group A2.26, Mechanical condition assessment of transformer winding using Frequency Response Analysis (FRA)
Dr.ing. Raka Levi
Based on notes from a lecture given in Melbourne, Australia April 2013