Ground Distance Relays Understanding the Various Methods of Residual Compensation, Setting the Resistive Reach of Polygon Characteristics,

, and Ways of Modeling and Testing the Relay

Quintin Verzosa, Jr. Jun Doble Engineering Company

Abstract
The correct operation of ground distance relays is highly dependent on the correct application of the residual or zero-sequence compensation factor. But what are these factors? Various relays vendors have different forms of naming, defining and applying these factors and this confuses relay application engineers. Names like KN, K0, KE, KG, KZN, RE/RL, XE/XL, Z0/Z1 and others are used; some factors are simple scalar numbers and others are in Vectorial form. A factor K0 in one relay can have a different definition from that of another relay that uses exactly the same name, and this may be true whether the relays come from different vendors or even from the same vendor. An incorrect compensation setting can result in either misoperation of the relay or its failure to operate for ground faults. This paper provides some derivations of residual compensation and a survey of the various forms of compensation, their definitions and how to convert from one form to another. Another confusion in the application ground distance relays that have polygon-shaped characteristics is the way the resistive reach is defined. In some relays the relay resistive reach setting is defined as a per phase resistance, which can be affected by the residual compensation factor; while in another relay it could be the ground loop resistance, which is not affected by the residual compensation factor. In addition the angle of the resistive blinder may or may not be influenced by the residual compensation angle. Hence, there is a need to accurately model and visualize the relay characteristic consistently, especially when comparing relays. In order to ensure that the relay is set correctly and operates as expected it is necessary to test the relay characteristic for ground faults and verify that it matches the relay settings and expected characteristics. Modeling the ground relay characteristic and using the correct residual compensation setting can be a complex subject for the test technician. Depending on how the relay resistive settings are entered, questions arise on how to model the relay polygon characteristics should it be modeled in the per phase plane, or in the loop plane, or should the reactance line be modeled per phase and the resistance line modeled per loop since many distance relays enter the settings as such? Even mho ground characteristics can be a challenge when the residual compensation setting is Vectorial with angles far away from zero. When testing the ground distance relay the capability of the protective relay test system becomes another important factor since the test system may or may not have the capability to exactly model the relay characteristic and residual compensation factor. One may have to convert the compensation to another form and model the characteristics in another way when testing and viewing the results of the test. This paper describes some available methods to do this and offers suggestions for testing ground distance relays. A good understanding of these concepts is essential for relaying personnel in order to properly set the relays and test them. Post mortem analysis of relay operations also requires a good understanding of these concepts.

Conventions used in this paper: Complex variables are shown in bold letters. For example Z is a complex impedance that consists of the resistance R and reactance X, or the magnitude Zmag and angle Zang, and may be written as: Z = R + jX = Zmag /Zang.

Ground Distance Relays Understanding the Various Methods of Residual Compensation, Setting the Resistive Reach of Polygon Characteristics, and Ways of Modeling and Testing the Relay

Residual Compensation and Zero-sequence Current Compensation

Most ground distance relays are designed so that they measure (and can be set) in terms of positivesequence impedance. Consider the system shown below where a relay 21G is required to measure the impedance, nZ1, from the relaying point to the phase A to ground fault at F. Quantities that are readily available or that can be derived include the relay voltage VaR, the current Ia, the residual current In, and zero-sequence current I0.
Es Z1S, Z0S

n
R Ia . Z1, Z0 F

VaR

21G

A-N Fault

Figure 1

In order to accomplish this goal to measure the distance to the fault in terms of positive sequence some form of compensation are used. Appendix A includes the derivation of some of these forms or techniques of compensation. In order to measure the fault impedance nZ1 ground distance relays modify the phase current Ia by adding to it a portion of the residual current In. One technique known as residual compensation adds a portion KN (called residual compensation factor) of the residual current In, to the phase current Ia. This modified current is used in the measuring circuit of the relay so that it measures the impedance as. Zrelay = VaR/(Ia + KNIn) = nZ1 Where KN = ( Z0/Z1 1) / 3 (1) (2)

Figure 2 shows a simplified network model equivalent to Figure 1, where it can be seen (see Appendix) that the ground return impedance is nZN = n(Z0-Z1)/3 and ZN = KNZ1
Relay Location EA Zs IB = 0 EB Zs n Z1
*

IA n Z1
*

VaR
EC Zs

I C =0 n Z1
*

Ph AN Fault IN Z Ns n ZN
*

=n (Z0 - Z1)/3
*

Figure 2 Simplified network model

Ground Distance Relays Understanding the Various Methods of Residual Compensation, Setting the Resistive Reach of Polygon Characteristics, and Ways of Modeling and Testing the Relay The concept of residual compensation is implemented in some relays as shown in Figure 3 where the phase current Ia (for a phase A to ground distance element) passes through a phase or positive-sequence replica impedance Z1 and the residual current In passes through a neutral replica impedance ZN. A residual compensation factor KN is used to adjust the value of ZN so that ZN = KN Z1. The net effect is the same as equation (1).
A B C A-N Fault IA

Van Ia
Relay Comparator Circuits Replica Circuits

Z1

Z1

Z1

ZN

In

Figure 3 An implementation of residual compensation in a ground distance relay

Another technique is zero-sequence current compensation. Here the phase current Ia is modified by adding to it a portion K0 (called zero-sequence current compensation factor) of the zero-sequence current I0. The relay measures the impedance

Zrelay = VaR/(Ia + K0I0) = nZ1

K0 = Z0/Z1 1

(3) (4)

Clearly both factors are related as K0 = 3KN, which is not surprising since In = 3I0.
In some relays the zero-sequence current compensation factor K0 is defined as in equation (2), which is actually the definition of KN; while in other relays K0 is defined as the ratio Z0/Z1, although internally

it may be used and substituted into equation (4).

Other compensation factor symbols used in some relays include KE, KG. And these could be defined using either equations (2) or (4). The factors KN, K0, KE or KG are also referred to as earth-return compensation factor, ground-return compensation factor, and neutral impedance correction factor. Other relays use only the reactance part of the impedances, X0/X1, and compensate for the angles of Z1 and ZN or Z0 but some relays make an approximation by assuming that the angles of Z1 and ZN are almost the same and that the reactance values are almost equal to the impedance magnitudes. In other relays the ratio RE/RL is used to compensate the resistance part and the ratio XE/XL is uses to compensate the reactance part separately. Yet, some relays are simply set in terms positive-sequence impedance Z1 (magnitude and angle) and zero-sequence impedance Z0 (magnitude and angle). And still other relays set the relay using R1, X1, R0 and X0. In both cases the residual compensation is automatically calculated internally by the relay using these set values. A list of formulas, compensation type names, and symbols, including other related variables to completely define the type of compensation, used in various distance relays is shown in Table 1. This is not a complete list of all terms used in all relays but it represents a majority of common relays. Other terms that are very similar are not included for brevity and clarity. Prefixes or suffixes that indicate the zone that the setting applies to have been removed.

Ground Distance Relays Understanding the Various Methods of Residual Compensation, Setting the Resistive Reach of Polygon Characteristics, and Ways of Modeling and Testing the Relay

Table 1 Compensation Formulas, Names and Symbols

Formula for calculating factors

KN magnitude and angle or K0 magnitude and angle or KE magnitude and angle or KG magnitude and angle or KZN res comp & KZN angle K0 magnitude and angle Z0/Z1 magnitude and angle K0 or ZerSeqK0 Z0/Z1 ratio KN KN Z1 angle ZN angle Z1 angle Z0 angle

This form of compensation is known as

Factor symbol or name (As set in the relay)

Other settings directly related to factor

Remarks

1 Z0 1 3 Z1

Residual compensation Zero-sequence current compensation Earth-return compensation Ground-return compensation Earth impedance compensation Neutral impedance correction

Zero-sequence current compensation

Z0 1 Z1 Z0 Z1

Zero-sequence current compensation

1 X0 1 3 X1

Residual compensation Neutral impedance correction

Relay internally calculates the correct complex factor using (|Z0/Z1| /angle) - 1 Relay internally calculates the correct complex factor using K0 /Z0angle Z1angle ZN angle=Z0angle-Z1angle Some relays use only the magnitude (scalar), angle=0. Z1 angle=ArcTan(tauK) ZN angle=ArcTan(tauN) Line L/R matching (ms) TauK=X1/R TauN=(X0-X1)/((R0-R1)) ZPh magnitude and angle ZN magnitude and angle KZPh , KZN, KZ1 Z1 angle, ZN angle Relay uses ZN directly in measurements Relay internally calculates the ZPh=KZ1KZPh=Z1 and ZN=KZ1KZN/KZPh

Residual compensation

Earth impedance (residual) compensation XE/XL

1 R0 1 3 R1 1 X0 1 3 X1
RE/RL and

Earth-return compensation Residual compensation Zero-sequence current compensation

R1, X1, R0, X0 or Z1 magnitude and angle Z0 magnitude and angle

Relay internally calculates the complex compensation factor = ( Z 0 / Z1 1)/3 where Z1=R1+jX1, Z0=R0+jX0

Ground Distance Relays Understanding the Various Methods of Residual Compensation, Setting the Resistive Reach of Polygon Characteristics, and Ways of Modeling and Testing the Relay

Conversion between Forms of Compensation

It is sometimes necessary to convert the values from one form of compensation to another. For example when replacing an existing relay one can use the existing settings and convert some setting as necessary. Of course, it is best to recalculate everything whenever possible. It may also be necessary to convert when testing since the relay may have some form of compensation that the test system does not support. In order to avoid ambiguity for the rest of the discussion, we shall use the following terms. These terms refer to the ground distance relay settings. It is possible that the positive-sequence ground element impedance settings may be different from that of the phase-phase elements both in magnitude and angle.
Z1 = positive-sequence impedance reach setting of the ground distance element Z1mag, Z1ang: magnitude and phase angle of Z1 R1, X1: resistance and reactance part of Z1 RL, XL: same as R1 and X1 Z0 = zero-sequence impedance (is used for calculating other compensation settings or is set directly) Z0mag, Z0ang: magnitude and phase angle of Z0 R0, X0: resistance and reactance part of Z0 ZN = ground- or earth-return impedance = (Z0 Z1) / 3 ZNmag, ZNang: magnitude and phase angle of ZN RN, XN: resistance and reactance part of ZN ZE = ZN RE, XL: same as RN and XN KN = residual compensation factor = ZN/Z1 =
1 Z0 1 3 Z1

KNmag, KNang: magnitude and phase angle of KN

K0 = zero-sequence current compensation factor =

Z0 1 Z1

K0mag, K0ang: magnitude and phase angle of K0 K0ratio = zero-sequence current compensation ratio = |Z0|/|Z1|

1 X0 1 3 X1 1 R0 RE/RL = earth-impedance compensation resistance ratio = 1 3 R1 1 X0 XE/XL = earth-impedance compensation reactance ratio = 1 3 X1
KNx = residual compensation factor using reactance part only=

Ground Distance Relays Understanding the Various Methods of Residual Compensation, Setting the Resistive Reach of Polygon Characteristics, and Ways of Modeling and Testing the Relay
ZLoop = Phase-to-ground loop impedance (excluding the arc resistance, which will be dealt with separately) = Z1 + ZN ZLoopMag, ZLoopAng: magnitude and phase angle of ZLoop RLoop, XLoop: resistance and reactance parts of ZLoop Zp = positive-sequence impedance of the fault point from the relay location Rp = positive-sequence resistance of the fault point from the relay location Xp = positive-sequence reactance of the fault point from the relay location

Table 2 includes a list of formulas on how to convert one form of compensation to another.
Table 2 Conversion from one form of compensation to another

From

To

Conversion 1

KNmag and KNang K0mag and K0ang K0ratio and Z0ang K0ratio and Z0ang K0mag and K0ang KNmag and KNang RE/RL and XE/XL KNmag and KNang
Z1 and ZN Z1 and Z0

K0mag and K0ang KNmag and KNang K0mag and K0ang KNmag and KNang K0ratio and Z0ang K0ratio and Z0ang KNmag and KNang RE/RL and XE/XL KNmag and KNang KNmag and KNang KNmag and KNang RE/RL and XE/XL

K0mag = KNmag3 K0ang = KNmag KNmag = K0mag/3 KNang = K0mag

K0 = K0ratio /(Z0ang-Z1ang) 1 K0 = (K0ratio /(Z0ang-Z1ang) `1)`/`3

K0ratio /Angle01 = K0 + 1 Z0ang = Angle01 + Z1Ang K0ratio /Angl301 = (3KN + 1) Z0ang = Angle01 + Z1Ang
KN = (RE/RL) cos(Z1ang) + j (XE/XL) sin(Z1ang) KNang = Im{KN} KNmag = Re {KN}

RE/RL=KNmagCos(KNang+Z1ang)/Cos(Z1ang) XE/XL=KNmagSin(KNang+Z1ang)/Sin(Z1ang)
KN = (ZN/Z1)

1 Z0 KN= 1 3 Z1
KN = [(R0+jX0)/(R1+jX1) 1)] / 3

R1, X1, R0, X0 R1, X1, R0, X0

RL/RL = XE/XL =

1 R0 1 3 R1 1 X0 1 3 X1

It is assumed that the characteristic angle Z1ang is known. Z1ang and set Z0ang are not necessarily equal to the line impedance angles.

Ground Distance Relays Understanding the Various Methods of Residual Compensation, Setting the Resistive Reach of Polygon Characteristics, and Ways of Modeling and Testing the Relay For fault points that lie on the characteristic setting angle Z1ang, all the above methods of compensation are equivalent and will have exactly the same measurement when applied correctly. These formulas have been implemented in a spreadsheet as shown in the images Figure 4. This spreadsheet allows one to select the type of compensation, enter the setting values, and perform the conversion. All the values in the tables are equivalent to one another. A graph of the phase or positive-sequence impedance, ground-return impedance and loop impedance together with a mho characteristic and the equivalent loop characteristic, helps in visualizing their relationships.

Then enter the setting values below:

Figure 4 An MS Excel implementation of conversion compensation formulas

Ground Distance Relays Understanding the Various Methods of Residual Compensation, Setting the Resistive Reach of Polygon Characteristics, and Ways of Modeling and Testing the Relay

Loop Impedance, ZLoop

Although the ground distance relay may be set in terms of positive-sequence or per-phase impedance, in reality the loop impedance is measured. Figure 2 (loop network) shows that relay voltage VaR is the voltage drop across the loop impedance Ia (Z1+ZN); or the voltage to current ratio is
VaR/Ia = Z1 + ZN = ZLoop

Figure 5 shows this relationship in the RX diagram, including their resistive and reactive components.
XLoop

ZN

XN

ZN = KN*Z1 = 1/3 (Z0/Z1 1)*Z1 K0 = (Z0/Z1 1) KN = K0 / 3 RL = R1 XL =X1 ZE = ZN RE/RL = 1/3 (R0/R1 1) XE/XL = 1/3 (X0/X1 1)

R1

ZNang

RN

ZLoop = Z1 + ZN = Z1 (1 + KN)

Z1

X1

ZLoopAng Z1ang

RLoop

Figure 5 Relationships of Phase, Ground-return and Loop Impedances

Converting to Loop Impedance. Given the fault point, Zp = Rp + j Xp in the positive-sequence impedance plane, Table 3 provides the formulas needed to calculate the impedances in the loop impedance plane. It is assumed that the characteristic angle Phi1 is always known for the factors that use angles.
Table 3 Formulas for Converting Fault Impedances along the Characteristic Angle to Loop Impedances Compensation type or Setting Type Other related settings ZpLoop RpLoop XpLoop

KN K0
K0ratio KNx Z0ang ZNang

ZpLoop=Zp (1+KN) ZpLoop=Zp (1+K0/3) ZpLoop=Zp [1+(K0ratio /(Z0ang-Z1ang)-1)/3 ] ZpLoop = RpLoop+jXpLoop

RpLoop=Re{ZLoop} RpLoop=Re{ZLoop} RpLoop=Re{ZLoop}

XpLoop=Im{ZLoop} XpLoop=Im{ZLoop} XpLoop=Im{ZLoop}

RpLoop=Rf+Xfptan(ZNang) XpLoop=Xp (1+KNx) RpLoop=Rp (1+KNx) RpLoop=Rp (1+RE/RL) RpLoop=Re{ZLoop} XpLoop=Xp (1+KNx) XpLoop =Xp (1+XE/XL) XpLoop=Im{ZLoop} XpLoop=Im{ZLoop}

no ZNang ZpLoop = RpLoop+jXpLoop RE/RL XE/XL ZpLoop = RpLoop+jXpLoop ZpLoop=Zp (1+ZN/Z1)

Z1 and ZN
R1, X1, R0, X0

ZpLoop=Zp{1+ RpLoop=Re{ZLoop} [(R0+jX0)/(R1+jX1)-1]/3}

Note: The table applies only to points that lie on the characteristic line angle setting. The exception is for those that use scalar compensation factors like KNx and RE/RL, XE/XL.

Ground Distance Relays Understanding the Various Methods of Residual Compensation, Setting the Resistive Reach of Polygon Characteristics, and Ways of Modeling and Testing the Relay Until now we have not yet included the fault resistance; this includes the arc resistance and the tower footing resistance, and is shown in Figure 6. The ratio V/I now becomes

Z1

AN fault

21

IaR
VaR

Rarc

VaR / Ia = Z1 + ZN + Rarc + Rtf = ZLoop + RFLoop

Rtf ZN

Figure 6
RLoop

Rarc

Rtf

Where ZLoop is the same as before. RFloop is the fault loop resistance and does not include the line resistance. Figure 7 shows the addition of the fault resistance plotted in the RX diagram. The RFLoop resistance is usually shown at the base of the RX diagram in relay manuals.

XLoop

RFLoop = Rarc + Rtg

ZN

XN

R1

ZNang

RN

ZLoopAng = Z1 + ZN

Z1

X1

ZLoopAng Z1ang

RFLoop

Figure 7

Characteristic Shapes and Resistive Reach

Figure 8 shows a survey of some of the characteristic shapes used in ground distance relay including the influence of the residual compensation on the resistive reach. Other shapes that are not shown exist but are not included here. The first shape (a) either use an impedance relay that is supervised by a directional element; shape (b) is a mho circular characteristic and shape (c) uses a reactance element supervised by a mho element. Two general terms used for shapes (d) through (j) are quadrilateral (four sides) and polygon (multiple angles). In this paper we shall refer to these shapes as polygons even though some have circular arcs. For the purpose of general comparison the shapes are drawn to have the same reactive reach (Z1 or X1) at the impedance set point along the protected line angle; and for shapes that have a specific resistive reach setting, they are set to measure the same loop resistance along the R-axis. The first three do not have a specific resistive reach setting. Shape (a) has less resistive coverage for shorter lines. Shape (b) mho characteristic has less resistive coverage especially for self-polarized elements; but for a cross-polarized mho element, it expands with increasing source impedance ratio (SIR), thereby increasing the resistive coverage. Shape (c) uses a reactance element that is supervised by a higher zone mho element and has more resistive coverage especially if the mho characteristic expands with higher SIR.

Ground Distance Relays Understanding the Various Methods of Residual Compensation, Setting the Resistive Reach of Polygon Characteristics, and Ways of Modeling and Testing the Relay

Z1

Z1

phi1

phi1

(a) Resistive reach depends on Z1 setting

(b)Mho. Resistive reach depends on Z1 setting. Expands with higher SIR

(c) Reactance type. Supervised by a mho element which may also expand

phi1

RFloop

Z1

Z1

phi1

Z1

phi1

RFloop

(d) Separate RFLoop reach setting. Resistive blinder in the loop impedance plane may tilt according to the KN angle.

(e) Separate RFLoop reach setting. Resistive blinder in the loop impedance plane may tilt according to the KN angle Left and right blinders are parallel.

X1 phi1

RFloop

(f) Bullet characteristic. Separate RFLoop reach setting. Resistive blinder in the loop impedance plane may tilt according to the KN angle

(g) Separate RFLoop reach setting. Resistive blinder in the loop impedance plane remains the same as the phase impedance with no effect of KN.

X1

RFph

(h) Separate RFLoop reach setting. Resistive blinder in the loop impedance plane tilts according to the KN angle but other vertical resistive blinders are not affected.

(i) Separate setting for RFph. RFloop =RFph*(1+KNx)

X1

RFph

X1

RFph

(j) Separate setting for RFph. RFloop =RFph*(1+KNx)

(k) Separate setting for RFph. RFloop =RFph*(1+RE/RL). Left and right blinders are parallel.

(l) Separate setting for RFph. RFloop =RFph*(1+K0/3). Resistive blinders can have different inclination angles

Note: Other compensation forms like K0ratio, (Z1 and ZN), and (R1, X1, R0, X0) are used but are not shown. These are usually equivalent to KN Figure 8 Samples of characteristic shapes and the relation of compensation and resistive reach settings

10

Ground Distance Relays Understanding the Various Methods of Residual Compensation, Setting the Resistive Reach of Polygon Characteristics, and Ways of Modeling and Testing the Relay The next four shapes (d) through (h) have a separate and specific resistance reach setting, which we shall call RFLoop in this paper. The term ohms per loop is usually used in some relays to indicate this fact. The resistive setting is the desired reach to cover the fault arc resistance Rarc and the tower footing resistance Rtf. (See Figure 6). An extra margin of 10% to 20% is usually added to cover more fault resistance. Thus, RFloop = (1.1 to 1.2) (Rarc + Rtf) The above RFLoop setting formula is valid for radial lines or when the infeed from the remote end to the fault is negligible; but this is not the case in most transmission line applications. For non-radial lines, the setting should be further increased by an additional ratio factor to handle remote infeed as shown in Figure 9.
Z1 AN fault Z2

21

IaR
VaR

Rarc

Iremote

Rtf ZN

Figure 9 Resistance measurement with fault resistance and remote infeed current

Thus, the setting is usually calculated as

RFloop = (1.1 to 1.2) (1 + Iremote/IaR) (Rarc + Rtf) The next four characteristics have a per-phase resistive reach setting RFph (in this paper) since the reach setting is calculated and referred to the positive-sequence part of the fault loop. In some relays the term, ohms per phase, is used to indicate this. The actual resistive reach is affected by the residual compensation factor. In calculating the resistive setting, the desired loop resistance coverage is divided by (1 plus the compensation factor). Most but not all relays of these types usually have a scalar compensation factor. For shapes (i) and (j), which use residual compensation KNx =(X0/X1-1)/3, the resistive setting is RFph = (1.1 to 1.2) (1 + Iremote/IaR) (Rarc + Rtf)/(1+KNx) For shape (k), which uses ratio compensation RE/RL, the resistive setting is RFph = (1.1 to 1.2) (1 + Iremote/IaR) (Rarc + Rtf)/(1+RE/RL). For shape (l), which uses residual compensation K0 = (Z0/Z1-1), the resistive setting is RFph = (1.1 to 1.2) (1 + Iremote/IaR) (Rarc + Rtf)/(1+ K0/3) Note that for faults on the line and assuming zero fault resistance the point will fall on the ZLoop line. Most numerical relays adjust the inclination of the resistive blinder to be the same as that of the angle of ZLoop, in order to maintain a constant resistive reach over the entire impedance reach setting. One thing that needs to be pointed out is that for some shapes the effective resistive reach may be less than the setting for faults away from the relay location. For shapes like (i) and (j) which have vertical

11

Ground Distance Relays Understanding the Various Methods of Residual Compensation, Setting the Resistive Reach of Polygon Characteristics, and Ways of Modeling and Testing the Relay resistive blinders, and shape (g) where the angle of the resistive blinder remains equal to that of the positive sequence impedance angle setting, it can be seen in Figures 10 and 11 that the resistive reach decreases.
Less RF coverage at end of line

RFph

RFLoop
Figure 10 Figure 11

The resistive reach may need to be increased further if needed. However, the maximum resistive reach that can be set on the relay is limited or affected by other factors relay maximum resistive setting, maximum load, use of load encroachment feature, relay current sensitivity, and the tilting effect of remote infeed current. These aspects are not covered in this paper.

Modeling & Testing

In order to properly test the relay one has to know the expected relay characteristic that conforms to the settings; we call this the reference characteristic. In this paper we shall focus our testing discussion mainly on the first quadrant of the R-X diagram for the following reasons: 1. A huge majority of faults fall in the first quadrant. 2. The settings and characteristics in the first quadrant are more well-defined and easier to model and test. 3. The directional lines affect the fourth quadrant as well as the second quadrant in most of the relays. For test methods that consider the source impedance the directional line moves down into the third quadrant. The characteristics and behavior of the directional lines varies to a high degree and can be the subject of another paper. Most test methods do not handle the directional lines very well. Although the focus as regards testing is on the first quadrant, some issues need to be considered when modeling the reference characteristics: 1. The form of residual compensation. 2. Relay impedance or reactance reach settings are in ohms per phase, i.e., positive-sequence values, but the resistive reach could be set either in ohms per loop or ohms per phase. One needs to be careful, as it may be difficult to find this information in some relay manuals. 3. The tilt or droop of the reactance characteristic affects the accuracy of some testing methods. 4. The inclination angles of the resistive blinders which may be fixed or set.

12

Ground Distance Relays Understanding the Various Methods of Residual Compensation, Setting the Resistive Reach of Polygon Characteristics, and Ways of Modeling and Testing the Relay 5. Whether or not the resistive blinder angle of inclination is affected by the residual compensation angle. 6. Based on the above information, should we model and test the relay using a per-phase impedance plane or a loop impedance plane? Before discussing the polygon characteristics, let us first consider the simpler mho characteristic shown in Figure 12 having a reach setting Z1mag Ohms/phase, characteristic angle Z1ang, and a residual compensation factor KN. The reference characteristic is usually modeled directly in the per-phase plane as shown by the smaller mho circle. The characteristic looks exactly as set. From equation (1) the impedance measured is Zrelay = VaR/(Ia+KNIn). But since In=Ia for a single line to ground fault and replacing Zrelay with ZFault, we can write the voltage equation as VaR = ZFault (Ia + KNIa) = Ia ZFault (1+KN). The ratio of the relay voltage to the current is then VaR/Ia = ZFault (1+KN). Any point on the diagram (Z1, P2, P3, P4) is a ZFault point on the per phase plane. Multiplying ZFault by (1+KN) gives the equivalent points (Z1L, P2L, P3L and P4L) in the per loop plane.
Per-phase model
Z1L

Constant test current method VaR = Ia ZFault (1+KN)

P2L

Loop Characteristic

Z1

Constant test voltage method Ia = VaR / (ZFault (1+KN))

P2 P3L

Per Phase Characteristic

P3 P4 P4L

Loop model

Constant test current method VaR = Ia ZFltLoop Constant test voltage method Ia = VaR / ZFltLoop.

Figure 12 Mho per-phase and loop characteristics

So given any point ZFault and the test current Ia, the test voltage VaR can be calculated. This is called constant test current method. VaR = Ia ZFault (1+KN) If the test voltage is given the test current can be calculated, this is called constant test voltage method Ia = VaR / (ZFault (1+KN)) When the relay is modeled per phase the KN factor needs to be considered in the calculation. The relay can also be modeled in per loop plane as shown by the larger circle. The loop model has a diameter O to Z1L at the ZLoopAngle. Since the effect of KN is already included in the loop model, it should not be included again in the calculation of the test voltages and currents. Each of the test points shown is tested using the equations VaR = Ia ZFltLoop and Ia = VaR / ZFltLoop.

13

Ground Distance Relays Understanding the Various Methods of Residual Compensation, Setting the Resistive Reach of Polygon Characteristics, and Ways of Modeling and Testing the Relay

The polygon characteristic shown in Figure 13 has setting X1 and RFph, both in Ohms/phase as shown in the smaller figure below. The residual compensation factor is a scalar KNx = (X0/X1-1)/3. Because the compensation factor is scalar (KN angle=0) it is even simpler than the mho characteristic. To get the loop impedance for testing, one only needs to multiply the per-phase fault point by (1+KNx)
P1L P2L P4L

P3L P1 P2 P3 X1 P5 RFph P6 P6L P4 P5L

Figure 13 Polygon characteristic with residual compensation KNx = (X0/X1 - 1) / 3

We shall now take the case where the resistance reach setting is in Ohms/loop while the reactance line setting is in Ohms/phase. Figure 14 on the right shows the how a typical ground distance zone characteristic would be depicted in the relay manual. Z1mag/Phi1 is the per-phase reach setting in Ohms/phase and RFLoop is the resistance setting in Ohms/loop. The residual compensation factor KNmag/KNang is another setting. In this relay the resistive reach is constant. Only the ground-return impedance is compensated.

Z1

Phi1
RFLoop

Figure 14 Polygon characteristic with RFLoop resistance setting and compensation factor KN
Ia
+ Va

Figure 15 on the right shows a simplified model (for a phase A-to-ground fault) of a ground distance element having a per-phase reach setting Z1 and residual compensation KN and a resistance setting RFLoop. The distance to the fault along the setting angle is represented by n. Note that the resistance RFLoop is not affected by the location of the fault.

n*Z1

RFLoop Ia n*Z1*KN

Figure 15 Equivalent circuit diagram

14

Ground Distance Relays Understanding the Various Methods of Residual Compensation, Setting the Resistive Reach of Polygon Characteristics, and Ways of Modeling and Testing the Relay If we consider the setting reach point Z1 in Figure 16 and multiply it by (1+KN) we get the loop impedance ZLoop, from the origin to Z1L. Now we can add the loop resistance RFLoop to ZLoop and we get the upper right-hand corner of the loop characteristic, Q. Drawing horizontal lines of length RFLoop starting from the ZLoop impedance line shows this constant resistive reach also in the loop impedance plane. Note that the resistive blinder is parallel to the ZLoop line, i.e., the angle of the resistive blinder in the loop characteristic is PhiLoop. This constant resistive reach in both per-phase and loop impedance planes allows us to treat the fault resistance separately from ZLoop. Consider point P2 in Figure 17 on the resistive blinder of the per-phase characteristic. What would be the equivalent point on the loop characteristic? We can draw a horizontal line from P2 until it interests the per-phase Z1 line at point P2. The length of this line is RF2. Now multiply the value of point P2 by (1+KN) and we get P2L, which lies on the ZLoop line. This is equivalent to drawing a line parallel to ZN (=KNZ1) until it intersects the ZLoop line. If we now add the resistance RF2 to P2L we get the point P2L which should lie on the loop characteristic resistive blinder. This is the loop impedance that we use for testing. We can do the same process for the other points P3 though P5 to get the loop impedances P3L through P5L. To test each point we substitute these loop impedances into the formula VaR = Ia ZLoop in order to calculate either the voltage or current. When testing such characteristic in the per-phase plane, a test method that treats the fault resistance separately has been implemented. In Figure 18 on the right ticking the checkbox +R direction will treat the fault resistance of all test points to the right of the characteristic angle as a separate resistance, and will not be affected by the KN factor. The R direction checkbox, when not checked, allows simultaneous testing a circular part of a bullet characteristic. .
Z1L RFLoop

Z1

PhiLoop

ZLo o

PhiLoop

RFLoop

Figure 16 Characteristic loop impedance with constant resistive reach

Z1L

ZN

RF2 P2L

P2L

Z1

RF3

P3L

P2
RF3

RF2 P3

P2

RF4 P4

P4L

RF4

RF5

P5 = P5L

Figure 17 Determining total fault loop impedance when modeling separate fault resistance.

Figure 18 Testing with separate fault resistance

15

Ground Distance Relays Understanding the Various Methods of Residual Compensation, Setting the Resistive Reach of Polygon Characteristics, and Ways of Modeling and Testing the Relay Figure 19 shows results of modeling and testing such a relay in both the per-phase plane as well as the loop impedance plane. The calculations to get the loop impedance models reference characteristic can be found in Appendix B

Per-phase test

Loop testing result with per-phase reference characteristic superimposed

Figure 19 Modeling and testing a polygon characteristic using separate fault resistance in the per-phase impedance plane compared with loop modeling and testing.

For characteristics having drooped or tilted reactance lines and with complex number compensation factors, testing modeling, testing with this method results in error for test points on the reactance line, especially for large angle tilt settings, as shown in Figure 20. Testing for a point P outside the reactance line will result in a point PL that is well inside the characteristic in the loop impedance plane. This results in the relay operating for a fault outside of the characteristic. For a small tilt angle less than 2 degrees, the error is hardly noticeable. For tilt angles larger than +5 degrees it is recommended to model and test the characteristic in the loop impedance plane. This is also one of the problems when testing the directional lines.

PL

ZLo op

RF

PL

RF

Rs1 Rs2

Figure 20 Error when using separate fault resistance with reactance lines having large tilt angles.

To model the characteristic of this type of distance element on the loop impedance plane, one has to draw the loop reactance line parallel to the original per-phase reactance line and passing through the ZLoop reach point. The blinders are drawn parallel to the ZLoop line and passing through the RLoop

16

Ground Distance Relays Understanding the Various Methods of Residual Compensation, Setting the Resistive Reach of Polygon Characteristics, and Ways of Modeling and Testing the Relay resistive setting points, Rs1 and Rs2 in this diagram. The angle of the directional line depends on the relay design. In some relays the angle remains the same as the per-phase characteristic, while in others it may change depending on the KN angle. If the angle of the residual compensation KN is equal to zero or very near to zero, it is also possible to test this characteristic in the per-phase plane without treating the arc resistance as separate. The new per-phase resistance is then RFph = |ZFLoop| / (1+KNmag). This is equivalent to modeling the relay using as scalar compensation KNx. The loop impedance used for calculation of voltages and currents would then use VaR/IaR = Zp(1+KN). This method of modeling and testing should not be used if the angle of KN is more than two degrees away from zero.

Some relays have fixed vertical blinders in the 2nd and 4th quadrants; i.e., they are not affected by the KN angle in the same way are the resistive blinder angle in the first quadrant is when calculating the loop characteristic. The left-hand resistive blinder may behave similarly but is not apparent when modeled in per phase but appears in the loop model as shown in Figure 21. This kind of characteristic, if tested only for the first quadrant, can be tested per phase with a separate arc resistance. However, if the entire characteristic needs to be tested, the only way to test it is to model the complete loop characteristic. Yet some relay characteristics have left-side blinders that are never allowed to enter into the first quadrant as shown in the Figure 22; the reactance axis becomes the limit. The first quadrant can be tested using a per-phase model; but if the left side needs to be tested, a loop impedance model should be used for the entire characteristic.

Figure 21 Modeling polygon loop characteristics with vertical blinders in the 2nd and 4th quadrants

Figure 22 Modeling polygon loop characteristics where the left blinder not allowed to enter into the 1st quadrant.

17

Z1

ZL oo

Ground Distance Relays Understanding the Various Methods of Residual Compensation, Setting the Resistive Reach of Polygon Characteristics, and Ways of Modeling and Testing the Relay For other ground distance elements that have complex number residual compensation factors and at the same time with characteristics shapes that have the resistance blinder angle different from the ZLoop angle it is difficult to test them in the per-phase plane. One would have to model these elements in the loop impedance plane. Figure 23 shows an example where the resistive blinder remains the same as the positive-sequence impedance or per-phase angle, phi1. There are also some relays that allow the user to set the resistive blinder angle values that those mentioned above.
TiltAngle

XN

Loop

phiN

TiltAngle Per-phase

X1
ZLoopAng

phi1

RFloop

phi1

Figure 23 Loop modeling of characteristic with resistive blinder that remains parallel to the phase impedance.

For relays that use compensation setting factors RE/RL and XE/XL the loop impedance characteristic is modeled simply by multiplying the resistive and reactive components separately. Thus, The resistive part is: PRLoop = PRPh (1 + RE/RL) The reactive part is: PXLoop = PXPh (1 + XE/XL) For most relays that use these constants, these equations apply to the entire characteristic, i.e. to all quadrants. The equations also apply to directional lines as well as tilted reactance lines. The problems described above in relation to Figures 20 through 23 do not usually exist here. The figure below shows the result of testing on the per-phase plane using compensation factors RE/RL and XE/XL. The figure on the right shows the loop characteristics and the result of testing using loop impedance, with the per-phase model (smaller one) superimposed. The loop calculations for this characteristic can be found in Appendix B.

Per-phase testing

Loop testing with per-phase characteristic superimposed

Figure 24 Modeling and testing characteristic that use compensation factors RE/RL and XE/XL

If the test software supports this form of residual compensation it is better and straightforward to simply model and test in the per-phase impedance plane and avoid unnecessary extra work of converting to loop impedance.

18

Ground Distance Relays Understanding the Various Methods of Residual Compensation, Setting the Resistive Reach of Polygon Characteristics, and Ways of Modeling and Testing the Relay A protection test system software with graphical interface and automated calculation of characteristics has been implemented that makes it easy for test personnel to model the relay by simply entering the settings. The equivalent reference characteristics are automatically modeled and plotted. In Figure 25 the following are entered: the type of characteristic, the characteristic angle, residual compensation type and values. Several modes or forms of residual compensation can be selected depending on what the relay is using. If it is not available one has to convert to a form that is available. Figure 26 shows an example of a graphical user interface where one can enter and define straight lines and circular arcs that comprise the characteristic.
Figure 25. Selecting the residual compensation form and entering relay parameters

Figure 26 Graphical and tabular interface for modeling and editing distance relay characteristics

Selecting a pre-defined shape allows direct entry of the settings, which then automatically adds the lines and arcs to the table. Examples of such interfaces are shown in Figure 27.

Figure 27 Pre-defined shapes for entering relay settings to automatically model relay characteristics

19

Ground Distance Relays Understanding the Various Methods of Residual Compensation, Setting the Resistive Reach of Polygon Characteristics, and Ways of Modeling and Testing the Relay Even with these modeling tools one has still to decide in some cases whether to model in per-phase or loop. If the characteristic and compensation form allows correct testing in the per-phase impedance plane, it is better to model it that way. Modeling in loop plane would, in many cases, require some auxiliary calculations to get the new reach points and angles. When in doubt, model and test using both loop as well as per phase impedance planes to see the effects of the compensation on the resistive blinder as well as the resistive reach. Appendix B includes worked examples on modeling polygon characteristics in the loop impedance plane. Some relay vendors export files setting that can be imported by the test system to automatically create the impedance functions for all zones of the distance relay. This eliminates the necessity to do manual calculations, especially when it needs to be modeled in the loop plane. Figure 28 shows such characteristics that have been imported.

Figure 28 Modeling distance relay zones

If one wants to test the polygon elements together with load encroachment elements that have circular arcs, modeling in the loop impedance plane will avoid the load circular arc part from looking like an elliptical arc. Figure 29 shows the result of simultaneous testing of all ground distance zones together with load encroachment. The directional lines and reverse zones are also tested.

Figure 29 Simultaneous testing multiple ground distance zones having load encroachment elements

20

Ground Distance Relays Understanding the Various Methods of Residual Compensation, Setting the Resistive Reach of Polygon Characteristics, and Ways of Modeling and Testing the Relay

Conclusions
1. Ground distance relays employ some form of compensation of the ground-return impedance in order to measure (and also to allow the relay to be set) in terms of positive-sequence impedance. The paper presents these forms of compensation including some derivation. 2. The many names, symbols and formulas that are in use for residual or ground-return compensation pose a challenge to personnel who set and test the relays. Moreover, some forms of compensation that use different formulas are called by the same name and symbol. This can result in applying the wrong setting if one is not careful and may result in either relay misoperation or failure to trip. 3. The fault resistance reach, for polygon-shaped characteristics, is set in ohms per phase in some relays while in other relays it is set in ohms per loop. In some relay manuals this fact is not explicitly indicated and relaying personnel will need to find this out. 4. The ground-return compensation affects the fault resistance reach and the angle of the resistive blinder in different ways, depending on the design of the relay. In some case the line remains vertical and in other cases it remains equal to the positive-sequence impedance angle. 5. Each form of ground-return impedance compensation can be converted to another form. Formulas are derived to perform this conversion. These formulas are handy when a relay being tested has a compensation setting that is not supported the relay test system. A spreadsheet that implements these formulas makes conversion easy and avoids calculation errors. 6. In addition to testing both reactance line and rsistance blinder in the per-phase impedance plane, and both in the loop impedance plane, a third test method that treats the fault resistance separately from the main loop impedance is described for testing polygon-shaped characteristics. The third method models the reactance line in per-phase and the resistive reach in Ohms per loop. 7. Selecting the most suitable model for testing depends on assessment of item 4 above as well as the tilt angle of the reactance line. Testing points for a reactance line that has a large tilt angle, using a separate fault resistance, will result in test errors. 8. Personnel who set relays and those who test them must have a good understanding of the methods of residual compensation, how the resistive reach is set and affected by the compensation and how the relay characteristics are modeled. Cooperation between these personnel is very important to actually verify their understanding of the settings and relay behavior and that the models are suitable. 9. The operation in the second and fourth quadrants of polygon characteristics is affected by additional factors including the behavior of the directional lines, the type of characteristic lines (straight lines of circular arcs), and the source impedance. 10. Some automated testing software support modeling the various forms of residual compensation setting and the resistive blinder so that the reactance line can be modeled in per-phase. If the test software does not support this the best way to test the relay is to model the entire polygon characteristic in the loop impedance plane and include the effect of residual compensation.

21

Ground Distance Relays Understanding the Various Methods of Residual Compensation, Setting the Resistive Reach of Polygon Characteristics, and Ways of Modeling and Testing the Relay

References
1. J. Lewis Blackburn, Protective Relaying Principles and Practice, New York and Basel: Marcel Dekker, Inc. 1987, pp. 436-438. 2. GEC Measurements, Protective Relays Application Guide, The General Electric Company p.l.c., St. Leonard Works, Stafford ST17 4LX, England, 1987, pp. 198-199. 3. S. E. Zocholl, Three-phase circuit Analysis and the Mysterious K0 Factor, 22nd Annual Western Protective Relay Conference, Spokane, Washington. October 24-26, 1995. 4. W. A. Elmore, Protective Relaying Theory and Applications, 2nd edition, New York, NY, Marcel Dekker, Inc. 2004, 254-256. 5. Anderson, P. M., Power System Protection, New York, The Institute of Electrical and Electronics Engineers, Inc., 1999, pp. 410-412. 6. Ungrad, H., Winkler, W., Wiszniewski, A., Protection Techniques in Electrical Energy Systems, New York, Marcel Dekker, Inc. 1995, pp. 134-137. 7. Horowitz, S. H., Phadke, A. G., Power System Relaying, 2nd edition, Taunton, Research Studies Press Ltd, Taunton, Somerset, England TA1 1HD. 1995, pp. 113-122. 8. Distance relay instruction manuals and reference manuals from various relay vendors. 9. F6TesT Visual Relay Testing software, Doble Engineering Company.

22

Ground Distance Relays Understanding the Various Methods of Residual Compensation, Setting the Resistive Reach of Polygon Characteristics, and Ways of Modeling and Testing the Relay

Appendix A. Derivation of ground-return compensation factors

Residual Compensation Factor, KN
Consider a system shown in Figure a1, with a ground distance relay at bus R protecting a transmission line, where a phase A-ground fault occurs at point F at a distance n from the relaying point. The distance n is in per unit of the relay setting Z1 and Z0.
n
R Z1S, Z0S Ia . Z1, Z0 F

Es

VaR

21G

A-N Fault

Figure a1

The symmetrical component sequence diagram for the system above for a single line-to-ground fault is shown in Figure a2.
Fault Location F, VF=0
I1 V1R
F1 F2 F0

I2 V2R

I0 V0R

nZ1 Relay Location R ZS1 Pos. Seq. Network

N1

nZ2 = nZ1

nZ0

ZS2 Neg. Seq. Network

N2

ZS0 Zero. Seq. Network

N0

E1

I1 = I2 = I0 .

Symmetrical Component Network for SLG fault at F

Figure a2

The voltage at the fault point F is zero, assuming that the fault is bolted, and the sequence voltages are: V1R = I1 nZ1 V2R = I2nZ1 V0R = I0nZ0 (a1) The phase A to neutral voltage VaR at the relay point is then the sum of the sequence voltages
VaR = V1R + V2R + V0R = I1 nZ1 + I2 nZ1 + I0 nZ0

(a2)

23

Ground Distance Relays Understanding the Various Methods of Residual Compensation, Setting the Resistive Reach of Polygon Characteristics, and Ways of Modeling and Testing the Relay The phase A current Ia at the relaying point is then
Ia = I1 + I2 + I0

(a3) (a4)

and, since I1 = I2 = I0, the residual (neutral) current is In = Ia + Ib + Ic = 3I0 I0 = In / 3 = Ia / 3

Adding and subtracting I0nZ1 in equation (a2), factoring out nZ1 and I0, and substituting equations (a3) and (a4)
VaR = = = = I1 nZ1 + I2 nZ1 + I0nZ1 I0nZ1 + I0nZ0 ( I1 + I 2 + I0 ) nZ1 I0nZ1 + I0 nZ0 Ia nZ1 + I0(Z0 Z1) n Ia nZ1 +(Ia/3)(Z0 Z1) n

(a5) (a6)

If we use the voltage VaR and the current Ia directly for measurement the apparent impedance that is measured is
ZRapparent = VaR / Ia = n Z1 + (Z0 Z1) n/3

The extra second term makes the result not very usable. To make the relay easier to use, the objective in the design of most ground distance relays is to make the relay measure only the first term, nZ1. If we substitute In/3 for I0 in equation (a5) and multiply the second term by Z1/Z1 and simplify the equation to express the impedances as a factor of Z1, we obtain
VaR = Ia nZ1 +(In/3)(Z0 Z1) n Z1/Z1 = [Ia + In(Z0 Z1)/(3Z1)] nZ1

(a7)

If we define a constant KN= (Z0 Z1)/(3Z1) which is valid for a given line configuration and ground resistivity, VaR simplifies to
VaR = (Ia + KNIn) n Z1

(a8)

We can now use a modified current (Ia + KNIn), in which a portion KN of the residual current In is added to the phase current, to measure the impedance.
Zrelay = VaR / (Ia + KNIn) = n Z1

(a9)

This technique of measurement to express the fault impedance in terms of the positive-sequence impedance, by adding to the phase current a portion of the residual current, is known as residual compensation. The constant KN is called the residual compensation factor, which is defined as
KN = ( Z0 Z1) / 3Z1 = ( Z0/Z1 1) / 3

(a10)

24

Ground Distance Relays Understanding the Various Methods of Residual Compensation, Setting the Resistive Reach of Polygon Characteristics, and Ways of Modeling and Testing the Relay

Ground-return Impedance, ZN
Considering equation (a7) and expressing the voltage drops in terms of Z1,
VaR = Ia n Z1 + Inn(Z0 Z1)/3

(a11)

This is the loop voltage from the relay terminal to the fault point and back, through a ground-return impedance nZN=n(Z0-Z1)/3, to the neutral of the relay location. Hence, we can model the network as shown below
Relay Location EA Zs IB = 0 EB Zs n Z1
*

IA n Z1
*

VaR
EC Zs

IC =0 n Z1
*

Ph AN Fault IN Z Ns n ZN
*

Network Model

=n (Z0 - Z1)/3
*

Figure a3. Simplified equivalent circuit The impedance ZN is called the ground-return (or residual) impedance and is defined as
ZN = (Z0 Z1)/3

(a12)
KN = ZN / Z1

Note also the relationship ZN = KN Z1 or

(a13)

Zero-sequence current Compensation Factor, K0

In equation (a7), if we replace In by 3I0 we get
VaR = [Ia + 3I0(Z0 Z1)/(3Z1)] n Z1 = [Ia + I0(Z0 Z1)/(Z1)] n Z1

We now define a constant K0 = (Z0 Z1)/(Z1) and simplify the equation to

VaR = (Ia + K0I0) n Z1

(a14)

Using the relay voltage VaR and the modified current (Ia + K0I0) yields an impedance measurement of
Zrelay = VaR / (Ia + K0I0) = n Z1

(a15)

25

Ground Distance Relays Understanding the Various Methods of Residual Compensation, Setting the Resistive Reach of Polygon Characteristics, and Ways of Modeling and Testing the Relay This technique of measurement to express the fault impedance in terms of the positive-sequence impedance, by adding to the phase current a portion of the zero-sequence current, is known as zerosequence compensation. The constant K0 is called the zero-sequence current compensation factor, which is defined as
K0 = ( Z0 Z1) / Z1 = Z0/Z1 1

(a16)

Residual Compensation Factors, RE/RL and XE/XL

Some relays compensate separately for the resistive part and for the reactive part. The loop impedance for a phase to ground fault consisting of Z1 and ZN is shown in vector form in Figure a3.
X

= ZE

ZN

XE

RE XLoop
ZL p oo

Z1 ZL =

XL=X1

R
RL=R1

RLoop

Figure a4 Loop Impedance vector diagram To make it consistent with nomenclature of relays that use this type of compensation, the subscripts L and E are used in place of 1 and N. Hence, the loop impedance is
ZLoop = ZL + ZE

We can express ZLoop into its resistive and reactive components

ZLoop = RLoop + j XLoop

And we know that ZL = Z1 = R1 + j X1 = RL + jXL If we expand equation (a12) into its resistive and reactive components,
ZE = ZN = (Z0 Z1) / 3 = (R0 R1)/3 + j (X0 X1)/3 = RE + j XE

So

RE = (R0 R1)/3 and XE = (X0 X1)/3

(a17)

26

Ground Distance Relays Understanding the Various Methods of Residual Compensation, Setting the Resistive Reach of Polygon Characteristics, and Ways of Modeling and Testing the Relay From Figure (a4) we see that the loop resistance is RLoop = RL + RE. If we express RLoop in terms of RL and introduce a constant RE/RL we get RLoop = RL + RE = RL (1 + RE/RL) If we now take the constant RE/RL substitute their respective formulas from equation (a17) RE/RL = [ (R0 R1)/3 ] / R1
RE 1 R 0 = 1 RL 3 R1

(a18)

Similarly, we can derive the reactive compensation factor XE/XL. Hence,

XE 1 X 0 = 1 XL 3 X 1

(a19)

Note the general inequality:

KN =

ZN ZE RE + jXE RE XE = = + j Z1 ZL RL + jXL RL XL

However, for any point along the loop impedance line the inequality does not exist.

27

Ground Distance Relays Understanding the Various Methods of Residual Compensation, Setting the Resistive Reach of Polygon Characteristics, and Ways of Modeling and Testing the Relay

Appendix B. Worked Examples in Modeling Some Polygon Characteristics in the Loop Impedance Plane
When the test software does not support testing, using the form of residual compensation and/or the resistive blinder is difficult to use for per-phase testing, it is necessary to model the ground distance polygon characteristic the loop impedance plane. This will simplify testing without bothering about these complexities in the calculation of test voltages and currents.

Example 1.
Given: The relay setting and characteristic is shown below. Reactive reach: Z1 = 5 Ohms @ 82 deg Resistive reach: 10 Ohms per loop Reactance line tilt: -3 deg Directional line angle: -8 deg Residual compensation: KN = 0.774 @ -14.29 deg
P1 TiltAngle
RBlinder

XPhLine
RBlinder

ZPh

ZphAng O -RFLoop

ZphAng DirAng

RFLoop

ZphAng

Figure b1 Per-phase impedance characteristic

Additional information: The residual compensation factor formula is: KN = (Z0-Z1)/3Z1 The resistive reach blinder on the left-hand side the same as that of the right-hand blinder The resistance blinder angle is influenced by the residual compensation The reactance reach tilt angle setting is that of the loop characteristic The angle of the directional line is affected by the residual compensation in the same way as the resistive blinder. Required: Model the loop characteristic, defining the line segments with a point and angle. Draw the loop characteristic. 1. Calculate the loop impedance ZLoop = ZPh /ZphAng * (1+KN) = 5 /82 * (1 + 0.774 /-14.29 ) = 8.80 / 75.77 28

Ground Distance Relays Understanding the Various Methods of Residual Compensation, Setting the Resistive Reach of Polygon Characteristics, and Ways of Modeling and Testing the Relay

2. The loop reactance line can be drawn. It passes through point P2 (which is the tip of the loop impedance) and has an angle 3 deg, which is the same as the original line. 3. The right-hand side resistive blinder passes through the point RFLoop (the resistive setting) and has an angle equal to the ZLoop impedance of ZLoopAng=75.77 deg. 4. The left-hand side resistive blinder passes through the point RFLoop, also with an angle equal to ZLoopAng=75.77 deg. 5. The directional line is affected be the residual compensation by the same angle as that of the blinders. The new directional line angle is equal to the original set angle plus the difference of ZloopAng and ZPhAng. DirLoopAng = (-8) + (75.77-82) = -14.23 6. The directional line is drawn passing through the origin and has an angle of 14.23

erLo op

XLoopLine

P2

TiltAngle
RBlin derLo op
RBlinder

RBlin d

XPhLine
RBlinder

ZPh

ZLoo p

P1

TiltAngle

ZLoopAng O -RFLoop

ZLoopAng RFLoop DirLoopAng

ZLoopAng

Figure b2 Loop impedance characteristic after applying KN compensation

29

Ground Distance Relays Understanding the Various Methods of Residual Compensation, Setting the Resistive Reach of Polygon Characteristics, and Ways of Modeling and Testing the Relay

Example 2.
Given: The relay setting and characteristic is shown below. Reactive reach: Z1 = 5 Ohms @ 82 deg Resistive reach: 3.107 Ohms per phase Reactance line tilt angle: -3 deg Directional line angle (right): -22 deg Directional line angle (left): 120 deg Residual compensation factors: RE/RL = 2.109, XE/XL = 0.723
P1 (R1, X1)
RBlinder

TiltAngle
RBlinder

XPhLine

ZP h

DirAng2 ZphAng - RFph ZphAng ZphAng RFph

DirAng1

Figure b3 Per-phase impedance characteristic with compensation RE/RL & XE/XL

Additional information: The reactive reach line is made up of a horizontal line to the left of the reach point P1 and a tilted line to the right of the reach point. The resistive reach blinder on the left-hand side the same as that of the right-hand blinder. It is also known that the relay directional lines are affected in the by the compensation factors in the same way as the reactance line and the blinders. Required: Model the loop characteristic, defining the line segments with a point and angle. Draw the loop characteristic. 1. Consider two per phase points P1 (R1, X1) and P2 (R2, X2) on the R-X diagram. The line connecting these two points has a slope of
Tan = X 2 X1 R 2 R1
P1' (R1', X1')

(b1)

P1 (R1, X1) P2 (R2, X2)

P2' (R2', X2')

Figure b4

30

Ground Distance Relays Understanding the Various Methods of Residual Compensation, Setting the Resistive Reach of Polygon Characteristics, and Ways of Modeling and Testing the Relay

2. One can calculate the loop points P1 and P2 using the compensation constants R1 = R1 (1+RE/RL) X1 = X1 (1+XE/XL) R2 = R2 (1+RE/RL) X2 = X2 (1+XE/XL) 3. The new slope of the per-loop line that connect these two points is
Tan ' = X 2' X 1' X 2(1 + XE / XL) X 1(1 + XE / XL) ( X 2 X 1) (1 + XE / XL) = = R 2' R1' R 2(1 + RE / RL) R1((1 + RE / RL) ( R 2 R1) (1 + RE / RL)

Substituting equation (b1),

Tan ' = Tan

1 + XE / XL 1 + RE / RL

The new line angle is 1 + XE / XL ' = ArcTan Tan 1 + RE / RL

(b2)

This formula will be used in calculating the new angles of all per loop line segments. 4. All reactance lines will pass through point P2, which is the ZLoop reach. The ZLoop reach components are: RLoop = ZPh*Cos (ZPhAng) * (1+RE/RL) = 5*Cos(82) * (1 + 2.109) = 2.163 Ohms XLoop = ZPh*Sin (ZPhAng) * (1+XE/XL) = 5*Sin(82) * (1 + 0.723) = 8.531 Ohms ZLoopMag = Sqrt(RLoop^2 + XLoop^2) = 8.801 Ohms ZLoopAng = ArcTan(XLoop/RLoop) = 75.77 deg 5. The loop line segment to the left of point P2 has an angle of 0 and the loop reactance line to the right of P2 has a new angle of TiltAngleLoop = ArcTan { Tan (-3) * (1 + 0.723) / (1 + 2.109) } = -1.664 deg

31

Ground Distance Relays Understanding the Various Methods of Residual Compensation, Setting the Resistive Reach of Polygon Characteristics, and Ways of Modeling and Testing the Relay

6. The resistive blinders have a reach of RLoop=RPh* (1+RE/RL) = 3.107* (1 + 2.109) = 10 Ohms 7. The resistive blinders now have a new inclination angle of RBlinderAngle = ArcTan { Tan (80) * (1 + 0.723) / (1 + 2.109) } = 75.77 deg It can be proven that this angle is actually equal to ZLoopAng. 8. The loop directional line angles are DirLoopAng1 = ArcTan { Tan (-22) * (1 + 0.723) / (1 + 2.109) } = -12.621 deg DirLoopAng2 = ArcTan { Tan (120) * (1 + 0.723) / (1 + 2.109) } = 136.172 deg The loop impedance model is drawn below, based on these calculated points and angles.
P2
op RBlin derL o

TiltAngleLoop XLoopLine
erLoo p

RBlinder

ZLoo p

P1

TiltAngle
RBlinder

XPhLine

ZP h

DirLoopAng2

ZLoopAng ZphAng - RFph -RFLoop

ZLoopAng
ZphAng

RFph

ZLoopAng

RFLoop

DirAng

DirLoopAng1

Figure b5. Loop impedance characteristic after compensating for RE/RL and XE/XL

32

RBlin d

Ground Distance Relays Understanding the Various Methods of Residual Compensation, Setting the Resistive Reach of Polygon Characteristics, and Ways of Modeling and Testing the Relay

Biographical Sketch
Quintin Verzosa, Jr. Jun, received his BSEE degree from Mapua Institute of Technology (Manila, Philippines) in 1976. He joined National Power Corporation (Philippines) in 1978 as a Relay Engineer and was promoted to Manager of Power System Analysis & Protection and Manager of Protection, Control & Communications Engineering Design. In 1993 Jun joined GEC ALSTHOM (now AREVA T&D) where he worked on relay applications and testing and later became Protection Systems Engineering Design Manager. Jun joined Doble Engineering Company in 1998 as a Senior Protection Engineer where he is involved in protection test systems marketing, training and support, research and development of protection models and testing algorithms, and the application of new technology. He is a member of the IEEE and is involved in PSRC working groups.

33