White paper

Efficient Method to Accelerate Resistance Measurement of

Transformer LV Winding

Edis Osmanbasic
DV Power

Introduction

Winding resistance measurements is a standard test routinely performed on all transformers.

The results accuracy is very important in order to get a correct condition assessment of
transformers. The measurement is performed using the Kelvin four wire direct current (dc)
method. Current through the circuit is established once the dc voltage is applied to the winding
under a test. This current creates the magnetic flux in the transformer magnetic core. This
process is slow due to the inductance of the core [L] that acts as a damper and slows down that
process. It is important to understand that the correct value of a winding resistance cannot be
measured until the current and the inductance become stable, since these values change at the
beginning of the measurement process.

Voltage on the winding terminals:

Winding resistance x current

Inductance x change of current (equal to zero
if current is stable)

Current x change of inductance

(equal to zero if transformer is saturated)

Transformer saturation

Transformer saturation will reduce the transformer inductance L, thus reducing the measurement
time. The time needed for transformer saturation depends on magneto-motive force (MMF).

1
The MMF is:

where “N” is the number of winding turns, and “I” is the current through the winding. According to
the equation above, it is easy to notice that magneto-motive force can be increased by
increasing the current value, or number of winding turns, or both.

Winding resistance measurements of LV (low voltage) transformer side may take a long time,
especially if the LV windings are connected in delta (triangle) configuration. The time needed for
the winding resistance measurement of large transformers can exceed 30 minutes per phase.
Delta winding connection, where the current flows through all three phases, causes the long
resistance stabilization time. Other reason for the long measurement time is a small number of
winding turns on the LV transformer side, requiring a very high current to reach the core
saturation.

As previously described the resistance measurement of LV windings can be accelerated by

saturating the core in one of the two ways:

- using a higher test current value I,

- increasing the number of turns N.

Higher test current value

To perform an accurate winding resistance measurement in a reasonable time, the transformer

should be saturated. The nominal current value through the transformer’s LV winding is usually
very high. It is necessary to perform the measurement with the minimum test current value of
1-5% of the transformer nominal current to achieve transformer saturation. For example,
generator step-up transformers may have nominal current values of 20 000 A on the LV side. In
most cases a portable test set cannot generate 1-5% of that current. Additionally, using the high
test currents requires thicker and heavier test cables.

Higher number of winding turns

Using the HV winding turns in the test current loop, as an additional help for transformer
saturation, will increase the total number of turns (Figure 1).

2
 Figure 1. Using HV and LV winding in series to increase the number of turns in the current loop

In this case the magneto-motive force will be:

where “N1” is the number of winding turns on the HV transformer side and “N2” is number of
winding turns on the LV transformer side.

The HV transformer winding contains significantly higher number of turns, compared to the LV
winding. Injecting the test current through the HV and LV winding will significantly increase
magneto-motive force, thus get the transformer into saturation faster and significantly reduce the
measurement time.

Case study

In this case study the winding resistance measurement of the LV transformer side is performed
in two ways:

- injecting a high current value through the LV winding only,

- injecting a lower current value through the HV and LV windings connected in series to
increase the number of turns.

Test object information:

Transformer type Generator step-up

Nominal power 400 MVA

Nominal voltage 420 / 20 kV

Nominal current 550 / 11533 A

Vector group YNd5

3
Test instruments:

DV Power Brand Comment Cables

Model 1 RMO100TT Single-phase portable device Thick / heavy

with max test current of 100 A
Model 2 TWA40D Three-phase portable device with Normal / light
max test current of 25 A per
phase

The first measurement was performed using the conventional method; connecting the test
device between two phases of the transformer LV windings. The intention was to use as high
test current value as possible (but not higher than 10-15% of the nominal current to avoid
heating of the winding). The test current value of 100 A was applied, which is about 1% of the
transformer LV rated current.

Figure 2. Diagram of LV winding measurement, green arrows represent magnetic flux

The figure 2 illustrates the test current flow and the flux directions in the magnetic core when the
winding resistance measurement is performed using a high current injected only through the LV
winding.

In the second measurement approach, the HV windings are used as an additional help to
achieve faster saturation. The connection may be established in such a way that the current
flows through the HV and the LV winding under test, located on the same transformer core leg.
Since the HV winding contains many more winding turns compared to the LV winding, the
magneto-motive force is increased significantly. The turns ratio of this particular transformer is

4
21 (420 kV/20 kV) hence the HV winding has 21 times more turns than the LV winding.
Optionally, for the HV wye (Y) connection the test current can be injected through all three
phases of the HV side (A, B, and C). The current source of the measuring device (the ⊕ output
of the instrument source) is connected to the phases A and C. The phase B (HV side) is
connected to the phase b (LV side) while the returning path is established through the phase a
(of the LV side) that is connected back to the source (the ⊖ output of the source). This way the
magnetic flux distribution through the core legs is established as presented in the Figure 3.

It is very important to establish the current through the HV and LV windings in such a way that it
creates the flux through corresponding transformer core leg in the same direction – boosting the
flux. The flux created from the HV side supports the flux created by the LV side and reduces
significantly the stabilization time (Figure 4).

This way, in case of YN configuration the complete magnetic core of the transformer is
saturated. The flux created by the current through the HV side will boost the flux created by the
LV side d (delta) in all three core legs.

Figure 3 Flux distribution in the transformer core during the normal transformer operation

5
Figure 4 Diagram of HV and LV winding turns connected in series to increase the number of turns in the current loop,
red arrows represent total flux flow

Measurement results

Test 1:

The test was performed according to the diagram in the Figure 2. The test current of 100 A was
used for the LV winding resistance measurement between transformer terminals a and b
(measuring Rab).The Figure 5 presents graph of the result stabilization process. The expected
resistance value was 1.135 mΩ. Ten minutes into the measurement process, the resistance
value has decreased to 1.165 mΩ, which still had not been the stable value. The resistance
value would have been slowly decreasing to 1.135 mΩ which would have lasted another 15
minutes or more.

6
 Figure 5 The winding resistance stabilization graph, the current flows through the LV winding only

Test 2:

This test approach was performed according to the diagram in the Figure 4. The test current of
25 A was injected (4 times lower current compared to the Test #1), but the HV side was used to
saturate the transformer. This way 21 times more winding turns of the same phase was applied,
plus a contribution of the other two phases with half of the current value each – equals about 42
times more ampere turns. In this approach the measured resistance value of 1.135 mΩ was
established after only 4 minutes (Figure 6). The graph in the Figures 6 and 5 are presented with
the same vertical axis scale to provide a better visual indication of the speed, and difference of
stabilization process time.

Based on the transformer vector group, the TWA test instrument automatically defines the
appropriate internal connection to establish a current through the HV and LV windings to get the
magnetic flux in correct direction through the corresponding transformer core leg. In addition,
only one-time test-cables set up is needed. It saves additional time of the test procedure by
internally connecting the HV phase B and LV phase b terminals.

7
 Figure 6 The winding resistance stabilization graph, the current flows through both, the HV and LV transformer
windings

Conclusion:

The transformer saturation using the HV winding connected in series with corresponding LV
winding speeds up the test process.

- Testing approach defined in the method #2 can be performed with a lower test current
value achieving transformer saturation.
- The lighter, smaller cross section test leads can be used.
- Significantly less time needed for the measurement, because of much faster result
stabilization time needed.

All this shows the benefit of using proposed test methodology for a transformer winding
resistance measurement, especially for testing the low voltage winding of generator step-up
transformers.