Rating CTs for Low Impedance Bus and

Machine Differential Applications

Stanley E. Zocholl
Schweitzer Engineering Laboratories, Inc.

Revised edition released October 2003

Previously presented at the

55th Annual Georgia Tech Protective Relaying Conference, May 2001,
54th Annual Conference for Protective Relay Engineers, April 2001,
V Simposio Iberoamericano Sobre Proteccion de Sistemas Electricos de
Potencia, November 2000,
and Southern African Power System Protection Conference, November 2000

Previous revised edition released April 2002

Originally presented at the

27th Annual Western Protective Relay Conference, October 2000
 RATING CTS FOR LOW IMPEDANCE BUS AND MACHINE
DIFFERENTIAL APPLICATIONS
Stanley E. Zocholl
Schweitzer Engineering Laboratories, Inc.
Pullman, WA USA

INTRODUCTION
Differential relays provide speed and sensitivity for clearing internal faults in critical bus and
generator applications. These relays are fast enough to operate during the offset portion of fault
currents where high current magnitudes together with high X/R ratios conspire to produce CT
saturation and false differential current. Such high fault current applications require full rated
current transformers (CTs) to prevent saturation and for security rely on the percentage
differential slope characteristic to override incidental false differential. The total scheme must
restrain from operating for the most severe external fault and yet have the sensitivity to operate on
the minimum fault. In existing schemes, the application engineer must live with legacy CT
ratings. Even in new installations, the engineer must contend with switchgear space limitations or
with CTs with ratings dictated by company standards rather than by specific application. The
application engineer must determine the percentage slope setting in relation to the degree of
saturation that provides the required security. This paper explores the relation between the
percentage slope characteristic and the degree of saturation produced by a range of CT ratings
encountered in utility and industrial applications. The paper will show the plots of the trajectories
of operating current versus the restraint characteristic for field applications, and establish settings
versus fault currents and CT ratings.

IL IR
Operate
M (Trip) k
IL IR IO
Restrain
IRES (No Trip)

Figure 1 Differential Protection Figure 2 Percentage Characteristic

PERCENTAGE SLOPE CHARACTERISTICS

Figure 1 shows the single-line diagram of a machine differential relay where the currents IL and IR
are shown in the polarity of an internal fault. As shown in Figure 2, in a percentage differential
relay the magnitude of the sum of the currents entering the zone must exceed a predetermined
percentage of a restraint quantity. The restraint quantity may have one of several definitions.
Two common characteristics are given in Equations (1) and (2) below:

IL + IR = k IL - IR (1)

IL + IR = k( IL + IR ) (2)

1
 The left hand side in each equation is the operate signal equal to the magnitude of the sum of the
currents. The right hand side in each equation is the restraint signal that is multiplied by the slope
k. The restraint signal in Equation (1) is the magnitude of the difference of the currents. In
Equation (2), the restraint signal is the sum of the magnitudes of the currents. Both characteris-
tics produce the relay characteristic shown in Figure 2. Some relays use the average magnitude
of the currents as the restraint. In this case, the slope k in Equation (2) is half the slope setting of
the relay.
There is no other purpose for the percentage differential characteristic than to override false
differential current caused by CT ratio error. Furthermore, since the CT ratio errors are less than
10 percent in the linear range, the only purpose for the percentage slope characteristic is to
override ratio error from saturation caused by the dc offset. What is the nature of the errors
caused by saturation and how significant are they?
Figure 3 and Figure 4 illustrate the severity of the CT saturation error. Figure 3 compares the
saturated secondary current with the true ratio sine wave current for a C200 2000:5 CT for a
10,667 A fault. When processed by a microprocessor relay, the digital filter removes the dc
component and all harmonics, leaving only the fundamental of the waveform as shown in Figure
4. In the saturated waveform, the later part of the half cycle of ratio current is missing since it
flows into the magnetizing impedance and not to the burden. As a result, the fundamental
extracted from the saturated waveform is reduced in magnitude and shifts leading in phase with
respect to the ratio current as shown in Figure 4. These are the signals processed by the relay, the
worst case being when the left hand set of CTs saturates and the right hand set does not.

Figure 3 Saturation Caused By DC Offset. Figure 4 Fundamental Waveforms Extracted

C200 2000:5 CT, 2.0 W Burden. Fault Current By Digital Filter
10667 Amp. X/R =14

ALPHA PLANE DIFFERENTIAL CHARACTERISTIC

The percentage differential characteristic can be visualized by expressing Equations (1) and (2) in
terms of the ratio [1]:

(I R / I L ) = a + jb (3)

2
Equation (1) written in terms of the ratio becomes:

IR I
1+ = k × 1- R (4)
IL IL

Substituting the complex number (a+jb) for the ratio and expanding gives the equation:

1+ k2
a2 + 2 2
a + 1 + b2 = 0 (5)
1- k
By completing the square, Equation (5) can be recognized as a circle with radius and center at:

2k
Radius = (6)
1- k2

1+ k2
Center = - + j× 0 (7)
1- k2

Consequently, the percentage slope characteristic is a circle enclosing the stability (no trip) region
in the complex plane called the alpha plane. Similarly, Equation (2) written in terms of the ratio
of currents becomes:

IR I
1+ = k × 1+ R (8)
IL IL

Substituting the complex number (a+jb) for the ratio and expanding gives the equation of a
characteristic resembling a cardioid that circumscribes the circle characteristic:

r2 +
2
1- k2
( )
cos (q) - k 2 × r + 1 = 0 (9)

where r2 = a2+b2 and a = r×cos(q). Figure 5 and Figure 6 show the plots in the alpha plane.

Figure 5 Circle Characteristic Figure 6 Characteristic Resembling a

Cardioid

3
DETERMINING THE SLOPE FOR A GIVEN DEGREE OF SATURATION
The alpha plane characteristic provides a convenient way of determining the slope setting in
relation to a given degree of CT saturation. The method is to plot the ratio of currents (IR/IL) in
the alpha plane for an external fault and then choose slope k such that all the points are enclosed
by the circle or cardioid shaped characteristic. The current samples are obtained from a Matlab
simulation program GUI_2CTS.m [2] that calculates current samples at 57.87 microsecond
intervals for 6-cycles. The program interface shown in Figure 7 lists the CT data.

Figure 7 Matlab Program Graphic Interface

The 1815 samples of each of the two currents are re-sampled by the 16-sample per cycle cosine
filter to obtain the fundamental phasor values. The ratios of the samples are then calculated. The
samples are plotted and appear as a cluster of points in the left complex plane for an external
fault. A slope k is then selected so that the plot of the characteristic encloses all of the samples.
The degree of saturation depends on the saturation voltage Vs across the burden impedance given
in the following equation [3]:

æ Xö
Vs = ç1 + ÷ × I f Z b (10)
è Rø

where If is the maximum fault current in per unit of the CT rating

Zb is the burden in per unit of the standard burden
X/R is the X/R ratio of primary fault current.
Our procedure will be to start with the left and right set of CTs rated at C800 and apply the
maximum fault current that just produces the threshold of saturation for a predetermined X/R
ratio. The value of Vs equals 20 in this case [3]. This value of fault current will be retained in all
subsequent cases. The saturation voltage Vs will then be doubled from the previous case by
lowering the ANSI voltage class of one set of CTs by one standard rating. In this way the

4
 saturation voltage will take on the values 20, 40, 80, and 160. Each simulation will produce 101
phasor samples that plot as a cluster of points in the alpha plane. In each case the value of k will
be selected such that the circle characteristics will enclose all the points. The values of k will
then be plotted as a function of the saturation voltage Vs. Each simulation was run for a 2000:5
CT with a lead and relay burden of 1 ohm total and a winding resistance of 1.0 ohm. The fault
current in each case is an offset 10667 amps where the X/R ratio is 14. The cases are plotted in
Figure 8 thru Figure 11 below. The darker points in each plot are the ratios (IR/IL). The lighter
points are the reciprocal ratios (IL/IR). Note that the k in each plot is in percent. Plots using the
cardioid shaped characteristic are shown in Figure 12 and Figure 13.

CASE PLOTS

Figure 8 Vs =20 Fault Current 10667 A Figure 9 Vs = 40 Fault Current 10667 A

Figure 10 Vs = 80 Fault Current 10667 A Figure 11 Vs = 160 Fault Current 10667 A

5
 Figure 12 Vs = 80 Fault Current 10667 A with Figure 13 Vs = 160 Fault Current 10667 A
Cardioid Shaped Characteristic with Cardioid Shaped Characteristic

SATURATION VOLTAGE
The relation between the slope setting and the degree of saturation is given by plotting the slope k
as a function of the saturation voltage. The curve with its empirical equation shown in Figure 14
determines the slope setting for a given degree of saturation. This equation is valid for Vs < 150.

80
k = 0.824 (Vs ) - 0.00242 (Vs )
2

60
Slope k

40

20

0
0 50 100 150
æ Xö
Vs = ç 1 + ÷ Ir Z b
è R ø

Figure 14 Slope k Versus CT Saturation Voltage

BUS AND MACHINE DIFFERENTIAL APPLICATIONS

In machine differential applications [4] there is no reason for the terminal-side CTs and the
neutral-side CTs to have a different ANSI voltage rating. In fact, it is impractical to size CTs to
avoid transient saturation in a generator differential because of high current magnitude and high

6
X/R ratios encountered. Consequently, the rule is to select the largest practical rating and match
the terminal- and neutral-side CTs so that in the presence of saturation both sets of CTs produce
the identical waveform and the ratio (IR/IL) remains at –1 during an external fault.

5333 A

5333 A

87B 87B

All CTs C400, 2000:5

10667 A

Figure 15 Multi-Restraint Bus Differential

Unlike the machine differential application, identical CTs in a multi-restraint bus differential
scheme do not guarantee the same waveform in all CTs. For example, the bus differential scheme
in Figure 15 shows the maximum external fault on a feeder. The CT on the faulted feeder carries
the total fault current of 10667 A. For an X/R ratio of 14, Vs is 40 and the fault current produces
saturation. However, the CTs of the transformer and the tie carry only half the total current.
Consequently, they act as one equivalent C800 CT with a saturation voltage Vs of 20 that
produces undistorted secondary current. These are the exact conditions of the case shown in
Figure 9 and the k of 30 is obtained by using Vs equal to 40 in the equation developed from the
plot in Figure 14. The alpha plane plot and a plot of the operate versus the restraint current for
the case are shown in Figure 16 and Figure 17.

Figure 16 Vs = 40, k = 30 Figure 17 Operate Signal Versus Restraint

7
THE EFFECT OF REMANENT FLUX ON THE SLOPE SETTING
A high offset fault current is usually interrupted in a few cycles. Consequently, the duration of
the fault current can be much shorter than the time constant of the primary circuit. The
interruption leaves a remanent flux in the core that is not affected by normal load current and that
can only be removed by demagnetization. A survey of 141 current transformers on a 230 kV
system is reported in Annex C of reference [4]. In this survey, the remanent flux ranged from 20
to 80 percent of the flux at the threshold of saturation.
The ANSI voltage rating of a CT is the voltage caused by 20 times the rated 5-ampere secondary
current flowing in the standard burden. Furthermore, the volt-time area of the rated voltage
represents the flux at the threshold of saturation [3]. Consequently, the remanent flux in percent
of the flux at the threshold of saturation reduces the voltage rating of the CT by that percentage.
For example, a C400 CT with 40 percent remanent flux has been effectively reduced to a C240
rating.
The effect of remanent flux can be included in the calculation of the saturation voltage Vs by
modifying the standard burden as follows:

VANSI æ % Re manence ö
Zstd = ç1 - ÷ (11)
100 è 100 ø

The calculation of k for the case shown in Figure 16 and Figure 17 for a C400 CT with 40 percent
remanent flux is as follows:
Maximum Fault current IF = 10667 amps
X/R ratio of primary circuit X/R = 14
CT primary current rating CTrating = 2000
CT burden in ohms Zburden = 2.0 ohms
Percent remanent flux %Rem = 40

VANSI æ % Re m ö 400 æ 40 ö
Standard Burden Zstd = ç1 - ÷= ç1 - ÷ = 2 .4
100 è 100 ø 100 è 100 ø

æ Xö IF Zburden 10667 2.0

Saturation voltage Vs = ç1 + ÷ = (1 + 14) = 67
è R ø CTrating Zstd 2000 2.4
2
Slope setting k = 0.824Vs - 0.00242Vs = 44

The calculation shows that the 40 percent remanent flux increases the saturation voltage from 40
to 67 and the secure slope from 30 to 44. The alpha plane plot and the plot of the operate versus
the restraint current for the case are shown in Figure 18 and Figure 19.

8
 Figure 18 Vs = 67, k = 44, 40 Percent Figure 19 Operate Signal Versus Restraint
Remanence with 40 Percent Remanence

CONCLUSIONS
1. The percentage slope characteristic defined by Equation (1) is a circle when plotted in the
complex alpha plane. The characteristic as defined in Equation (2) is a characteristic that
resembles a cardioid that circumscribes the circle. The stability (no trip) region is the area
enclosed within the circle or cardioid-shaped characteristic.
2. The ratio of the differential currents (IR/IL) that occur during an external fault forms a
collection of points in the alpha plane that should reside within the circle characteristic.
3. The radius and center of the circle characteristic are determined by setting the slope k.
Consequently, the alpha plane characteristic provides a convenient way of determining the
slope setting in relation to a given degree of CT saturation.
4. It is impractical to size CTs to avoid transient saturation in a generator differential because of
the high current magnitude and high X/R ratios encountered. Select terminal- and neutral-
side CTs of identical rating and manufacture type.
5. The security of bus differential applications relies on the percentage slope setting. A bus
scheme with more than one source contributing to an external fault can be analyzed as a two-
CT equivalent with k determined by the CT saturation voltage Vs in the relation:

k = 0.824 × Vs - 0.00242 × (Vs )

2
(11)

6. The remanent flux expressed as a percent of the flux at the threshold of saturation reduces the
CT voltage rating by that percent. Consequently, the effect of remanent flux can be
incorporated into the calculation of the saturation voltage Vs. The remanent flux causes a
significant increase in Vs and in the secure slope setting.

9
REFERENCES
[1] Protective Relays: Their Theory and Practice - Volume I. A. R. van C. Warrington, 2nd ED,
1968, London, Chapman & Hall, pages 99-140.
[2] S. E. Zocholl and D. W. Smaha, “Current Transformer Concepts,” Proceedings of the
46th Annual Georgia Tech Protective Relay Conference, Atlanta, GA, April 29–May 1, 1992
Pages 7-9.
[3] S. E. Zocholl, G. Benmouyal, and Jeff Roberts, “Selecting CTs to Optimize Relay
Performance,” 23rd Annual Western Protective Relay Conference, Spokane, WA,
October 15–17, 1996, pages 2-4.
[4] IEEE Std C37.110-1996, IEEE Guide for the Application of Current Transformer Used for
Relaying Purposes (ANSI), Clause 7.2.2.6, pages 34; Annex C, page 55.

ACKNOWLEDGEMENT
The author wishes to thank Dr. Hector Altuve for introducing us to the classical attributes of the
alpha plane [1] and for his derivation of the alpha plane characteristics of percentage differential
relays.

BIOGRAPHY
Stanley (Stan) Zocholl has a B.S. and M.S. in Electrical Engineering from Drexel University.
He is an IEEE Life Fellow and a member of the Power Engineering Society and the Industrial
Application Society. He is also a member of the Power System Relaying Committee and past
chair of the Relay Input Sources Subcommittee. He joined Schweitzer Engineering Laboratories
in 1991 in the position of Distinguished Engineer. He was with ABB Power T&D Company
Allentown (formerly ITE, Gould, BBC) since 1947 where he held various engineering positions
including Director of Protection Technology.
His biography appears in Who’s Who in America. He holds over a dozen patents associated with
power system protection using solid state and microprocessor technology and is the author of
numerous IEEE and Protective Relay Conference papers. He received the Best Paper Award of
the 1988 Petroleum and Chemical Industry Conference and the Power System Relaying
Committee’s Distinguished Service Award in 1991.

Copyright © SEL 2000, 2002, 2003

(All rights reserved)
Printed in USA
20031009

10