Error Ratio Evaluation of Protection Core of Selected

Older Distribution Current Transformers

David Lazecky, Vladimir Kral, Stanislav Rusek, Zdenek Hytka

Faculty of Electrical Engineering and Computer Science
VŠB - Technical University of Ostrava
Ostrava, Czech Republic
david.lazecky.st@vsb.cz, vladimir.kral@vsb.cz, stanislav.rusek@vsb.cz, zdenek.hytka@vsb.cz

Abstract—Current transformer (CT) is one of the vital devices transformers which were available from local distribution
for metering electrical parameters of distribution and network operator. The need for such an evaluation was based
transmission networks. Requirements and specification are on non-availability of such experiment under equivalent
described in technical standards for 60 years at least, but ageing conditions, such as network configuration, electrical
and performance after some time of on-site operation have not parameters, climate etc. Moreover, the age and used-state, in
been described in recent works well, especially in conditions general, were considered as a benefit because of the intended
similar to network and climate situation in the Czech Republic. application of the data.
Due to the intended aim of using output values for technical state
estimation of network components in RCM strategy The main aim of the measurement is to evaluate the error
development, only protection cores measurement for error ratio ratio of a current transformer, provide corrective curves for
evaluation was proposed. The results of two selected older oil- evaluated transformers (if needed) and to investigate the
insulated current transformers (5P10 class) from distribution behaviour of the transformers in overcurrent and fault-current
network are presented in the paper. The cores are measured areas for the first harmonic in general. These information were
according to the protection CT accuracy class, and also, considered highly important for further research and
matching measuring CT accuracy class for a general overview of application of Reliability Centered Maintenance (RCM), such
CT accuracy only. The equipment used for measurement is as the projection of ageing and wearing of network
described. The highest current recordings, ratio error evaluation components.
and values are presented and depicted. Results applicability is
discussed.
A. Current Transformer
Keywords—Current transformer, distribution network Current transformer (CT) is a device used for continuous
equipment, error ratio, experimental evaluation, protection core interpretation of primary-side current in values that metering
devices connected to secondary-side of the CT can handle. It is
I. INTRODUCTION designed to be connected in primary circuit, the most common
connection is in series. Common values of secondary-side
Instrument transformers are essential parts of many
currents are 1 A and 5 A (nominal value), while the primary
metering, control and protection systems. The instrument
current is a subject of design need. In power electrical CTs,
transformer quality directly affects the accuracy and
there are two types of cores, one for measurement and the other
performance of these systems since they can be as accurate as
for protection, which are described below in accordance with
the provided network information. Two basic functions are
technical standards specifications. As in any real device, there
provided by the transformers: change of magnitude (but not the
are factors affecting CT performance and output values. In case
nature) of primary to secondary values; and insulation between
of CTs (and proper installation), they are connected with an
primary and secondary circuit. The majority of currently used
iron core, precisely its design. Two most important are
instrument transformers is still based on magnetic core design
described in the following paragraphs. [1, 2, 3]
although newly designed instrument sensors are available on
the market. [1, 2, 3] Saturation is one of the factors. It is problem caused by
magnetic flux in the CT core (magnetic capacity). The CT core
The requirements for instrument transformers are specified
saturates in case of any large-magnitude current. The output
in (national) technical standards. The predecessor for currently
values do not match the primary circuit current anymore and
valid standard is IEC 60044. For current transformers, IEC
they are (much) higher than the nominal current value.
60185 (and IEC 185) are even older. Currently valid standard
However, protection relays are designed to withstand the
family is IEC 61869.
situation and able to take action. The main problem comes
when the real magnitude of current in such a situation is
II. EVALUATION SPECIFICATION needed. [1, 2, 3]
The evaluation was designed as an experimental Presence of the dc component presence in the primary
measurement of on-site used, and therefore aged, instrument current is the other highly important problem of CTs. It usually

This research was partially supported by the SGS grant from VSB - Technical
University of Ostrava (No. SP2019/20) and by the project TUCENET
(No. LO1404).

978-1-7281-1334-0/19/$31.00 ©2019 IEEE

appears on an ac power system as a result of a transient event network asset fleet, some of the most common CTs in the
or fault. DC component high enough with a time constant long network.
enough offsets ac current and therefore a CT may reach the
saturation limit in some time and CT’s performance collapses. All CTs are oil-insulated type for outdoor usage with three
When the dc component disappears, the CT output returns back cores, one measurement core and two protection cores. Both
to normal operation state. [1, 2, 3] CTs are designed for up to 123 kV voltage level. The
manufacturer of each CT was insignificant for the selection.
The selection itself was done based on electrical network
B. Measurement Core operator consideration.
For measuring purposes, the CT core is designed to operate
under normal network state. Therefore, the primary current is The table below specifies the unit values most important for
expected to be between 0 % and 120 % of CT’s nominal the measurement. The rest of the CTs’ specification is not
current. Due to the billing use, the accuracy class is expected to listed due to its insignificance. Both CTs were configured for
be high in terms of current magnitude and angle. The main 600 A primary current.
accuracy classes and error values are depicted in the following
table. TABLE III. MEASURED CTS SIGNIFICANT VALUES

Measured core
TABLE I. ACCURACY REQUIREMENTS FOR CT MEASURING CORE [5] Mfg. IPRI
Mfg. Model ISEC Accur. Load
year (A)
Current error (±%) Angle error (±minutes) (A) Class (VA)
Accur. at % of In at % of In
class ABB IMB 123 1997 150-300-600 1 5P10 30
5 20 100 120 5 20 100 120
PFIFFNER JOF 123 1994 150-300-600 1 5P10 30
0.1 0.4 0.2 0.1 0.1 15 8 5 5
0.2 0.75 0.35 0.2 0.2 30 15 10 10
0.5 1.5 0.75 0.5 0.5 90 45 30 30 III. EVALUATION REALIZATION
1 3 1.5 1 1 180 90 60 60 The measurement of CTs was realized in the Centre for
Research and Utilization of Renewable Energy
(CVVOZEPowerLab), Brno University of Technology, Czech
C. Protection Core Republic via an open access policy. In addition, the
cooperation and technical support from the owner of the CTs
For protection purposes, the CT core has to be designed to
were provided too.
operate under various network disturbances (over currents and
fault currents). The values of normal operation are not so
important. Therefore only ratio error and phase displacement at A. Testing Circuit
rated primary current are specified, and composite error at The equipment of the high power lab was used. List of the
rated accuracy limit of primary current is introduced. The most important components follows. The high-power source
performance above the accuracy limit is not guaranteed. The consists of alternator-generator assembly. The generated power
accuracy classes and error values are depicted in the following is then led via switch to the first transformer. After the first
table. transformer, the main switch, for switching the circuit on at 0
degrees electrical, follows. The next component in the testing
circuit is the main testing transformer (delta/grounded Y
TABLE II. ACCURACY REQUIREMENTS FOR CT PROTECTION CORE [5]
connection). Tested CTs were connected to two phase-winding
Current Angle error outputs using appropriate cables.
Accur. Total error (±%)
error (±%) (±minutes)
class at accuracy limit The measured protection core (second core in both cases)
at In at In
5P 1 60 5
output was connected to a load consisting of two parallel-
connected resistors of 39 Ω and 4 A each. The two not-
10P 3 - 10 measured outputs were short-circuited.
Primary current values were measured on cables connecting
the source circuit and CT. The secondary-side measurement
D. Tested Current Transformers was performed on the cable connecting the CT output and the
The current transformers for the measurement were load.
obtained from one of the Czech electrical distribution
companies. The company selected them on the basis of B. Measuring Devices
availability of their spare part units. According to the provided
Values of the primary-current channels were gathered by
information, the measured CTs were installed and run under
using two measuring devices. Channel I.rc.4 (or I.rc.40) used a
common operating conditions in the past. In some time prior to
converter of the same name manufactured by company Elcom
the measurement, the CTs were, for non-specified reason,
Praha. The range is switchable between 4 kA and 40 kA.
replaced and stored as spare parts in case of another operated
Channel U.10.1 used a Rogowski coil and a converter module
unit failure. They should present a sample of the distribution
U.10.1 DU volt 10 manufactured also by Elcom Praha
company. The range of Rogowski coil is up to 100 kA per V, A. Data Processing
and; the resolution of the converter is 10 V. All measured datasets were imported to and processed
The secondary current values were gathered by current using Microsoft Office 2013 Excel software. Used procedure
transducer IT 200-S with resolution 200 A, which is a product description follows. All values are considered accurate and
of LEM company. As the primary side converter, the U.10.2 final.
DU volt 10 unit, manufactured by Elcom Praha company, was  Datasets were imported to MS Excel. Transient current
used. One of the cables connecting the CT and the load was component and dc offset were identified and omitted or
winded ten or five times around the transducer to increase the suppressed. Only stable section of each current was
transducer resolution capability according to the maximum used for further evaluation.
secondary current.
 All important values for the calculation and
IV. RESULTS measurement information, such as minimum,
maximum, RMS etc., were determined for all channels.
Measured values were gathered using LabView software
and saved in a proprietary format as files with *.lvm extension.  CT error ratio is calculated using the following
The time sampling rate is 0.02 ms for all measurements. For equations:
values between 10 % and 120 % of In, the source switch on
time is 600 ms. For the higher currents, the source switch on
time is 100 ms.  kI = IPRIn / ISECn 

where kI is CT ratio, IPRIn is the rated primary current,

and ISECn is the rated secondary current.

 I = ( ISEC  IPRIn – IPRI ) / IPRI 

where I is CT ratio error, ISEC is the measured

secondary current, and IPRI is the measured primary
current.
 For each measurement, the ratio error values are
collected and displayed in the tables below. Also, the
average from values for both channels is calculated.

B. Data Evaluation
All results are compared with limits specified in section II.
Fig. 1. The highest recorded currents, ABB CT Limits of the error ratio are specified in IEC 61689-2 technical
standard (or withdrawn IEC 185 as specified on one of the
CTs). For the evaluation of currents between 10 % and 120 %
of rated current, a compromise of using matching accuracy
class 1 is made although the protection cores were measured.
Moreover, the limit specified for accuracy class 5P is used for
higher fault-current values although the maximum specified
value is 10 In.
As mentioned before, measurement core full range
evaluation was done only for a general overview of CT
performance in the low current area. Accuracy class 1 is used
for specifying limits due to the same value of limit at CT’s
rated current. In value’s line is highlighted in the table, only
those values are used for the final evaluation of CT. While
ABB CT is most often within the measurement core limits, on
contrary, the Pfiffner CT is mostly outside, even at In.

Fig. 2. The highest recorded currents, Pfiffner CT

The highest recorded currents figures show the saturation

limits for both CTs. From the lower figure, it could be seen that
the Pfiffner CT has a higher limit than the ABB CT.
 TABLE IV. RATIO ERROR FOR UP TO 120 % IN TABLE V. RATIO ERROR FOR FAULT CURRENTS (10 IN AND MORE)

ABB PFIFFNER ABB PFIFFNER

I I
I.rc.4 U.10.1 avg. I.rc.4 U.10.1 avg. I.rc.40 U.10.1 avg. I.rc.40 U.10.1 avg.
(A) (%) (%) (%) (%) (%) (%) (%) (kA) (%) (%) (%) (%) (%) (%) (%)

60 10 0.77 0.97 0.87 1.7 1.48 1.59 6 1000 1.19 1.3 1.25 1.31 0.92 1.12
120 20 0.18 0.48 0.33 1.43 1.41 1.42 10 1667 1.2 1.38 1.29 1.24 0.85 1.04
180 30 -0.31 0 -0.15 1.47 1.38 1.43 12 2000 1.05 0.68 0.86
240 40 -0.33 -0.05 -0.19 1.42 1.32 1.37 15 2500 not measured 1.09 0.73 0.91
300 50 -0.33 -0.03 -0.18 1.26 1.16 1.21 18 2833 1.1 0.8 0.95
360 60 0.54 0.83 0.69 1.47 1.33 1.4 20 3333 -15.93 -15.75 -15.84 0.35 0.1 0.22
420 70 0.52 0.84 0.68 1.46 1.32 1.39
480 80 0.45 0.76 0.6 1.42 1.28 1.35 The measured values show that even two times higher
540 90 0.53 0.86 0.69 1.37 1.24 1.31 current than the accuracy limit is within the limits (using linear
approximation for ABB CT). For currents over accuracy limit,
600 100 0.46 0.77 0.61 1.36 1.22 1.29 the Pfiffner CT had much better performance than the ABB
660 110 0.46 0.78 0.62 1.3 1.18 1.24 CT. All values are inside specified limits. The limits’ lines
maximum is the value of current two times higher than the
720 120 0.56 1.82 1.19 1.28 1.64 1.46
accuracy limit, which was considered high enough.

Fig. 3. Current error ratio of both CTs’ protection cores for 10 – 120 % In, Fig. 4. Current error ratio of both CTs’ protection cores for 10 In – 20 kA,
average values (measurement core accuracy class 1 used) average values

The most important results of the measurement follow. The

following table shows the values of over-current values for
protection core limits. The first value is the rated accuracy
current, which is 6 kA for both CTs. Other values up to 20 kA
follow. The current of 20 kA value was considered as highest
possible at a substation and highest needed for the evaluation.
After this value, only more significant effect of saturation
could be expected. The Pfiffner CT was measured last, and due
to the availability of time reserved for the measurement, some
current values between 10 and 20 kA were added.
 V. CONLUSION
The paper presents the evaluation of two older current
transformers obtained from one of the Czech distribution
electrical grid operators. For some unspecified period of their
lifetime, the CTs were normally operated on-site, and at the
time of the measurement, they were listed and stored as backup
units. As described before, the purpose of the measurement is
based on a need of knowing the state of health and behaviour
of currently installed and operated CTs.
From the results, one could see that ABB CT passed the test
and standard-specified protection core limits are met. When
using measurement core limits, some values are outside.
According to the over-current measurements, the last measured
value above the current limit is definitely out of the core
designed parameters. On the other hand, the Pfiffner CT is not
performing well in the standard-current values and the nominal
current value is outside the limit, as same as the majority of the
results in the section. But the over-current values meet the
standard requirements in all measured values. Therefore, one
could deduce that the protection core was designed better and
for higher currents than the ABB core.
The results show the behaviour of ageing CTs. Without
knowledge of the previous degradation, they still provide an
important outcome for assumptions for RCM implementation
in an electrical grid. Based on the measurement, although only
two older operated CTs were measured, it could be presumed
that over-current values outside the CT specification could be
interpreted unbiased if a proper technique is used. Further
research will be aimed towards proper interpretation of over-
current values.

ACKNOWLEDGMENT
This research work has been carried out with the support of
CVVOZEPowerLab Research Infrastructure (project
No. LM2015092).

REFERENCES
[1] P. Gill, Electrical Power Equipment Maintenance and Testing, 2nd ed.
Boca Raton: CRC Press, 2009.
[2] W. Elmore, Protective Relaying Theory and Applications, 2nd ed. New
York: Marcel Dekker, 2004.
[3] C. Bayliss, B. Hardy, Transmission and Distribution Electrical
Engineering, 4th ed. Oxford: Elsevier, 2012
[4] IEC 60044: Instrument Transformers (family of standards)
[5] IEC 61869-2: Instrument transformers - Part 2: Additional requirements
for current transformers
[6] P. Mathews, Protective Current Transformers and Circuits. London:
Chapman & Hall, 1955.
[7] J. Král, J. Vaněk: Elektromagnetické přístroje, Brno: VUT, 1984.