A New Blocking Scheme

8/3/2019 A New Blocking Scheme

1/18

Volume 11,Issue 1 2010 Article 8

International Journal of Emerging

Electric Power Systems

A New Blocking Scheme for Distance

Protection during Power Swings

Ahmad Farid Abidin, Universiti Kebangsaan Malaysia

Azah Mohamed, Universiti Kebangsaan Malaysia

Hussain Shareef, Universiti Kebangsaan Malaysia

Recommended Citation:

Abidin, Ahmad Farid; Mohamed, Azah; and Shareef, Hussain (2010) "A New Blocking Scheme

for Distance Protection during Power Swings,"International Journal of Emerging Electric

Power Systems : Vol. 11: Iss. 1, Article 8.

DOI: 10.2202/1553-779X.2398

Available at: http://www.bepress.com/ijeeps/vol11/iss1/art8

2010 Berkeley Electronic Press. All rights reserved.


2/18

A New Blocking Scheme for Distance

Protection during Power Swings

Ahmad Farid Abidin, Azah Mohamed, and Hussain Shareef

Abstract

This paper presents a new and fast algorithm to prevent distance relay mal-operation during

power swings. The algorithm blocks the relay tripping signals during power swing and unblocks

the signals if a fault occurs during power swing. The proposed blocking scheme for distance

protection incorporates an additional criterion into the conventional relay which is based on thederivative of the line reactive power as seen by the relay. This technique overcomes the

shortcoming of conventional power swing detector (PSD) by removing the pre-defined R-X

diagram. The conventional PSD has the difficulty in obtaining the timer setting at pre-defined R-X

diagram due to varying cycle of power swings. To illustrate the effectiveness of the proposed

algorithm, simulations were carried out on the IEEE 39 bus test system using the PSS/E software.

Test results show the effectiveness of the proposed scheme in blocking the relays false trip

signals during power swing.

KEYWORDS: distance relay, fault, power swing, PSD, line reactive power


3/18

Introduction

The distance relay is one of the main protection components that have been usedto protect the transmission lines. Distance relay operates on the basis of

impedance at the point of measurement. The voltage and current are usually been

used in order to obtain the measured impedance. During a fault, the measuredimpedance is very low and it can enter the relay setting to initiate tripping signals

of associated breakers (Kundur, 1993; Ziegler, 1999; Horowitz, 2008). However,

the relay also may send trip signals during power swing due to low measuredimpedance during such circumstances. Power swings occur following a system

disturbance such as load changes and fault clearance. When a power swing

occurs, a change appears in the relative phase angle between two groups of

generators. As consequences, the measured voltage, current, apparent impedance,active power, reactive power and angle vary due to oscillations during power

swing. However, during a fault, these quantities change more significantly and

hence are frequently used to distinguish between a fault and a power swing toavoid relays false tripping signals.

Many techniques based on aforementioned quantities have been

introduced by a number of researchers (Jonsson and Daadler, 2001; Shaohua etal., 2005; Su et al., 2005; Xiangning et al., 2006; Lin et al, 2008; Zadeh and Li,

2008). Jonsson and Dalder, (2001) have introduced the application of negative

sequence current magnitude and the derivative of current angle to dictate the relayoperation during fault and power swings. This technique is very fast by sending

the blocking signals in 10ms according to the test conducted on the Nordic 32

lines system. Although the results were promising in blocking false trip signals

during power swing, the possibility of false trip signals during fault clearance wasnoted. The combination of waveform of swing centers voltage (WSCV) and

synthetic negative sequence vector (SNSV) has been utilized to differentiate

between a fault and a power swing (Shaohua et al., 2005). The technique seems tobe rigorous in discriminating power swing and high fault resistance for protection

purposes. However, it requires two computationally heavy steps of derivative

operation for WSCV. There is also a time delay of about 30-40 ms before a faultand a swing can be identified and hence the method is relatively slow as

compared to the technique of ref. 4.

A technique based on Vcos has been introduced by Su et al. (2005). Thetechnique takes about 30-50 ms to activate a power swing detector. However,

further testing is needed in larger power systems before the existing technique can

be deployed to the relay. The derivative of real power and reactive power hasbeen integrated to develop an unblocking scheme for distance protection during

symmetrical faults in power systems (Xiangning et al., 2006; Lin et al, 2008).

This technique is very complicated and computationally inefficient since it

1

Abidin et al.: A New Blocking Scheme for Distance Protection during Power Swings

Published by Berkeley Electronic Press, 2010


4/18

requires instantaneous product of voltage, current and angle to obtain the real andreactive powers. The unblocking scheme sends the trip signals after 30 ms in the

event of a fault. A more advanced technique using adaptive neuro fuzzy system

has been developed to block the relay trip signals during power swings (Zadehand Li, 2008). However, no justification has been done on Zone 3 relay operation

setting considering that this zone is the most vulnerable zone during power

swings. In addition, the relay response time is more than 40 ms which is very slowas compared to techniques developed by Jonsson and Daadler (2001), Shaohua et

al.(2005), Su et al. (2005), Xiangning et al. (2006) and Lin et al. (2008).

In this paper, a new and fast algorithm by using the derivative of the line

reactive power has been proposed to block false tripping signals during powerswings. In the algorithm, a new criteria based on the abrupt change in the

derivative of line reactive power during a fault is incorporated. In order to validate

the robustness of the technique, the algorithm has been tested on the IEEE 39 bus

system. A comparison of the proposed algorithm with the technique of ref. 4 isalso made to ascertain the validity of the proposed algorithm.

Formulation of New Criteria during Power Swing

The fundamental behavior of the line reactive power immediately before and after

a three phase fault can be explained by using a simple power system shown Fig. 1.

Eg

Vs Vload

Zg Zline=Rline+jXline

Zload=Rload+jXloadGenerator

Fig. 1 A simple power system

2

International Journal of Emerging Electric Power Systems, Vol. 11 [2010], Iss. 1, Art. 8

http://www.bepress.com/ijeeps/vol11/iss1/art8

DOI: 10.2202/1553-779X.2398


5/18

The proposed criterion is based on the fact that most of the reactive power duringa fault is consumed by the line reactance. The derivation of the criteria is shown

below;

||

(1)

By using the Kirchoff Voltage Law (KVL), the voltage at bus Vs can be writtenas;

(2)

Substituting (2) into (1), equation (1) becomes;

||

(3)

The the line impedance, Zline can be represented as;

Where the resistive component, RLine is very small and hence it can be neglected.

Thus, equation (3) becomes;

||

(4)

where,

Ss: Apparent power at sending end

Vs: Nominal voltage at sending endIs: Current flow at the line

ZLine: Impedance of line

RLine: Resistance of the line

XLine: Reactance of the line

ZLoad: Impedance of the loadVLine: Voltage at line

VLoad: Voltage at load

3




6/18

From Fig. 1, the apparent power of the system is composed of the combinedapparent power at transmission line and load which is given by;

(5)where,

SLoad: Apparent power of the loadPLoad: Real power of the load

QLoad: Reactive power of the load

SLine: Apparent power of the line

PLine: Real power of the lineQLine: Reactive power of the line

At transmission lines, reactive power is very large as compared to real power,

thus, we can assume that;

|| | |

Simplifying equation (5), we get,

(6)

During a power swing, the load impedance is significantly larger than the line

impedance. This relation can be formulated as;

(7)

Thus, by assuming and , equation (1) can be furthersimplified as;

||

(8)

Using similar assumption as in equation (7), we can assume that the reactive

power of the line, QLine is;

0 (9)

Unlike the case of power swing, the line impedance during a fault is very large

compared to the load impedance and . Thus, equation (1) becomes,

||

(10)

4



DOI: 10.2202/1553-779X.2398


7/18

This proves that the line reactive power change significantly during a fault.Equations (9) and (10) show that the line reactive power is abruptly changing

from 0 to|

|. This feature can be used as a criterion for distance relay operationto avoid triggering of false tripping signal due to power swing. Based on thiscriterion, it is possible to propose the use of rate of change of the line reactive

power, dQline/dt to discriminate between a fault and a power swing so that the

relay operates only due to a fault.

The Proposed Algorithm

The flowchart of the proposed algorithm used to prevent distance relay mal-

operation during a power swing is shown in Fig. 2. Normally, the relay

continuously monitors the apparent impedance, Za at a measured bus. Once the

apparent impedance enters the zone 3 relay, the relay will distinguish between afault and a power swing by using the rate of change of the line reactive power,

dQline/dt. The change of line reactive power during a fault is very substantial andit may produce very large positive value of dQline/dt compared to a power swing.

The fault identifying conditions of the algorithm, based on dQline/dt threshold of

fault (THfault) is described as follows;

A fault has occurred (11)

No fault has occurred (12)

In this study, the threshold value of fault (THfault) is set to 2000 Mvar/s.The next stage of the algorithm is to identify the fault clearing operation

by primary protection. When a fault is cleared, the dQline/dt will have a negative

value with large magnitude. The formulation of such operation is written as:

Fault clearance operation (13)

In this case, the threshold of fault clearance (THfault clear) is set to -2000 Mvar/s. If

a relay senses that the dQline/dt is below the threshold THfault clear value, the relaywill be reset due to the fault clearance operation although the apparent impedance

is at the relay boundary. Otherwise, the relay timer continues to monitor the

apparent impedance whichever falls at the relay zone. The relay will send atripping signal if the apparent impedance remains in zone 3 after the timer setting

is exceeded (normally 1-1.2 second).

5




8/18

Fig. 2 The flowchart of the proposed algorithm

6



DOI: 10.2202/1553-779X.2398


9/18

Simulation Results

The proposed algorithm is tested on the IEEE 39 bus test system by using thecommercial PSS/E software version 31. The test system consists of 10 generators

and 18 loads as shown in Fig. 3.

Fig. 3 The IEEE 39 bust test system

Five different fault cases have been considered to generate the powerswing condition as described below:

Case 1: Three phase fault at line between buses 5-8 from 1 to 1.15s, followed by

fault clearance and line tripCase 2: Three phase fault at line between buses 6-7 from 1 to 1.15s, followed by

fault clearance and line trip.

Case 3: Three phase fault at bus 5 from 1 to 1.15s, followed by fault clearance andbus disconnect

7




10/18

Case 4: Three phase fault at bus 6 from 1 to 1.15s, followed by fault clearance andbus disconnect

Case 5: Three phase fault at bus 11 from 1 to 1.15s, followed by fault clearance

and bus disconnectFrom the above cases, 3 different relays have been identified to be

operating falsely during power swings. The identified mal-operating relays on the

basis of case studies are:Case 1: Relay at bus 6

Case 2: Relay at bus 5

Case 3: Relay at bus 6

Case 4: Relay at bus 14Case 5: Relay at bus 14

In order to justify the reliability of the proposed algorithm, another three phase

faults need to be simulated during power swing occurrences. The locations of the

entire faults are simulated at 200% of the distance relay protected zone. Theseentire faults are simulated at 3-3.05 second of the simulation time. Bus 6 active

power variation during simulation is shown in Fig.4.

Fig. 4 Active power profile at bus 6

From Fig. 4, it is clearly observed that the first fault has been created at 1 second

until 1.15 second which causes a power swing to appear after the fault clearance.Subsequently, a second fault which is located at 200% away of the relay boundary

has been created at 3 second. During the simulation, the apparent impedance, Za

8



DOI: 10.2202/1553-779X.2398


11/18

seen by distance relay at bus 6 is very low during the fault and power swingconditions. The apparent impedance or impedance trajectory enters the relay

operating zone at both situations as shown in Fig. 5.

Fig. 5 The apparent impedance or impedance trajectoryenters the relay operating zoneduring fault and power swing at relay bus 6

Once the apparent impedance, Za enters the relay operating zone, the distancerelay may send the trip signals to the breaker to clear the fault. However, during

power swing, the trip signals should be blocked to avoid false tripping. The

proposed additional criteria can be introduced in distance relay in order to avoidsuch undesirable relay operation.

Results ofld/dtl

One of the fast techniques to discriminate a fault and a power swing employs the

use of negative sequence current magnitude and the magnitude of derivative of

current angle, ld/dtl (Jonsson and Daadler, 2001). However, identical values of

ld/dtl may appear during fault and fault clearance as the increment of current

angle is very substantial in both situations. The results in Fig. 6a and 6b showclearly the ld/dtl values of the affected relay during fault, fault clearance andpower swing for each case study described before.

9




12/18

Fig. 6a Result from ld/dtl

Fig. 6b Result from ld/dtl (enlarge)

0

2000

4000

6000

8000

10000

12000

case1 case2 case3 case4 case5

degree/second

Fault

Faultclearence

PowerSwing

0

10

20

30

40

50

60

70

80

90

100

case1 case

2 case

3 case

4 case

5

degree/second

Fault

Faultclearence

PowerSwing

10



DOI: 10.2202/1553-779X.2398


13/18

As can be seen from Fig. 6a, the range of ld/dtl is approximately between 2,100

degree/second to 9,500 degree/ second during the fault. The results of ld/dtl

values during power swing are depicted in Fig. 6b, where the value of ld/dtl is

between 9 degree/second to 45 degree/second. The results from both figures prove

that the ld/dtl is very promising in distinguishing between a fault and a power

swing. However, it can be noted that the ld/dtl values are not suitable to

differentiate between fault and fault clearance as shown in Fig. 6a. From Fig. 6a,

it can be deduced that, the value of ld/dtl is apparently in similar range for faultand fault clearance. The distance relay installed in these lines may send false trip

signals during fault and fault clearance operation as shown in Fig.7.

Fig. 7 Relay tripping signals while using ld/dtlat bus 6Further testing has been made for each effected relays to validate the

reliability of the ld/dtl approach as tabulated in Table 1.

11




14/18

Table 1: Relay tripping signals using ld/dtl

Case Fault Fault clearance Power

Swing

case 1 trip trip block



case 4 trip block block


From Table 1, it can be concluded that this technique is able to block the

relay trip signals during power swing. However, this technique is proven to be

vulnerable during fault clearance in which, 3 of the relays send the false trip

signals to the breaker since the ld/dtl is identical with fault ld/dtl values.

Therefore, an alternative approach needs to be introduced to avoid such adverseoperation.

Results of dQline/dt

The dQline/dt is used as a new indicator in order to enhance the capability of the

relay to differentiate between a fault, power swing and fault clearance in powersystems. Simulation results are displayed by plotting the Qline and dQline/dt

profile at the effected line against time as shown in Fig. 8 and Fig. 9, respectively.

12



DOI: 10.2202/1553-779X.2398


15/18

Fig. 8 Qline at at line between buses 6-11 which is associated to case 1

Fig. 9 dQline/dt at the line between buses 6-11 which is associated to case 1

The results in Fig. 10a and Fig. 10b shows that the magnitude of dQline/dt

for the effected relay during fault, fault clearance and power swing for each relay

that has been installed at the associated lines.

13




16/18

Fig. 10a dQline/dt for relay at associated lines during fault, fault clearance and powerswing

Fig. 10b dQline/dt for relay at associated lines during fault, fault clearance and power

swing (enlarge)

From the figures, it is clear that the range of dQline/dt for each case is

significantly different during fault; fault clearance and power swing. The range of

dQline/dt is between 60000 to 120000 Mvar/second, -90000 to -51000Mvar/second and 20 to 320 Mvar/second for fault, fault clearance and power

swing, respectively. The proposed technique has been tested on each relays forfurther justification. The results of each effected relays whilst using the technique

have been tabulated in Table 2.

150000

100000

50000

0

50000

100000

150000

case1 case2 case3 case4 case5Mvar/second

Fault

Faultclearence

PowerSwing

0

50100

150

200

250

300

350

400

case1 case2 case3 case4 case5

Mvar/second

Fault

Faultclearence

PowerSwing

14



DOI: 10.2202/1553-779X.2398


17/18

Table 2: Relay tripping signal using dQline/dt

Case Fault Fault clearance Power

Swing






From Table 2, it can be noted that the all the trip signals has been blockedduring fault clearance and power swing. Fig. 11 shows the relay trip signal while

using the proposed technique.

Fig. 11 Relay tripping signals at bus 6 while using dQline/dt of line between buses 6-11

It is clear that, the relay trip signals only appear during the fault. Theresults prove that the proposed technique is able to improve the reliability of the

relay operation as compared to the technique which has been proposed by Jonsson

and Daadler, 2001.

15




18/18

Summary and Concluding Remarks

The use of dQline/dt has been proposed as a technique to block the distance relaytrip signals during power swing. Time domain simulations were first carried out

under the conditions of fault and power swing. The proposed technique has been

applied and tested to evaluate its effectiveness in blocking the trip signals during power swing and fault clearance. The results show that the dQline/dt can

effectively block the trip signals during fault clearance and power swing unlike

the use ofld/dtl.

References

Horowitz , Stanley H. and Padke, Arun G. (2008) Power System Relaying, John

Wiley & Sons, Ltd, West Sussex, England.

Jiao, Shaohua ; Bo, Zhiqian; Liu, Wanshu ; Yang, Qixun (2001), New Principles

to Detect Faults During Power Swing, Seventh IEE International

Conference on Developments in Power System Protection, pp.515 - 518 .

Kundur, P., (1993)Power System Stability and Control McGraw-Hill, New York.Jonsson, M., Daalder, J (2001) A New Protection Scheme to Prevent Mal-Trips

Due to Power Swings, IEEE/PES Transmission and Distribution

Conference and Exposition: Vol. 2, pp. 724 729.Lin, Xiangning,; Gao, Yan; Liu, Pei (2008) A Fast Unblocking Scheme for

Distance Protection to Identify Symmetrical Fault Occurring DuringPower Swings, IEEE Transaction on Power Delivery: Vol 1, pp. 73-77.

Su, B.; Dong, X.Z.; Bo,Z.Q.; Sun,Y.Z.; Caunce, B.R.J.; Tholomier,D.; Apostolov,

A. (2005) Fast Detector of Symmetrical Fault During Power Swing for

Distance Relay, IEEE Power Engineering Society. General Meeting: Vol.

2, pp.1836 1841.

Xiangning, Lin; Qing, Zou; Wenjun, Lu; Kecheng, Wu; Hanli, Weng (2006) A

Fast Unblocking Scheme for Distance Protection to Identify SymmetricalFault Occurring During Power Swings, IEEE Power Engineering Society

General Meeting, pp. ,1-8.

Zadeh, Hassan Khorashadi; Li, Zuyi (2008) A Novel Power Swing Blocking

Scheme Using Adaptive Neuro-Fuzzy Inference System, Electrical Power

System Research. Journal, vol. 78, pp. 1138-1146.Ziegler, Gerhard (1999) Numerical Distance Protection Principles and

Application, Publicis MCD, Munich and Erlangen.

.

16



DOI: 10.2202/1553-779X.2398

A New Blocking Scheme

Documents