A NEW DIRECTIONAL OVER CURRENT RELAYING SCHEME FOR DISTRIBUTION FEEDERS IN THE PRESENCE OF DG

Transcription

1 A NEW DIRECTIONAL OVER CURRENT RELAYING SCHEME FOR DISTRIBUTION FEEDERS IN THE PRESENCE OF DG CHAPTER INTRODUCTION In plain radial feeders, the non-directional relays are used as they operate when the CT secondary current exceeds the threshold value of pickup setting in relays. This type of relay operates irrespective of the direction of current flow. The feeders other than plain radial feeders are not protected by the non-directional overcurrent relays as they require the creation of zones. The protection of such parallel feeders or double-end-fed feeders is protected by the directional relays. By introducing the directional feature in relays, uninterrupted supply can be made possible at all load points connected in parallel/ring system. In plain radial feeder as shown in Figure 3.1, if the breaker 1 trips because of any abnormalities in the section between bus A and bus B, it will interrupt the power supply at the buses B, C, and D. Thus, because of the tripping of former breaker, the load connected to the lateral buses will not receive power supply. A B C D Load R 1 R 3 R 5 Load Load Load Figure 3.1 Single line diagram of a radial system In case the same radial feeder is fed from both the ends (double-end fed) with necessary modification in the protection scheme using directional feature at relay point R 2, R 3, R 4, and R 5, as shown in Figure 3.2, zones are created for different sections between any two buses. 41

2 A B C D Gen-1 R 1 R 2 R 3 R 4 R 5 R 6 Gen-2 Figure 3.2 Double-end feed radial feeder In the event of any abnormalities (fault) in section between bus A and bus B, the breakers 1 and 2 will isolate the faulty section, without interrupting the supply to load the connected at the buses A, B, C, and D. Hence, to discriminate the faulty section, the relay R 2 should be direction sensitive so that it operates only in the direction indicated by the arrows as shown in Figure 2. It can be concluded that the relays R 2, R 3, R 4, and R 5 should operate for a current that flows away from the bus where the relay is located, and it restrains if the current flows towards the bus Directional Relay Characteristics The directional relay is a two input quantities relay that receives line current and bus voltage. The relay compares the direction of the current flow with reference to the bus voltage by measuring the phase angle between line current and bus voltage. The directional relay operates on watt metric principle, where the voltage coils (VCs) receive voltage from the bus potential transformer (PT) and the current coils (CC) receive the current from the line CT secondary. A maximum positive torque is produced when the current and voltage supplied to the CCs and VCs are in phase. Hence, the angle between current and voltage at which the relay develop maximum torque is defined as the maximum torque angle (MTA). In a directional relay, if V is the voltage given to the VC of the directional relay, then a current I V lags the voltage V by very large angle θ (because of inductive nature of VC). The flux produced by this current I V is Ф V. I is the current given to the current coil of directional relay, which setups a flux Ф I because of current I. If the angle γ between these two fluxes Ф V and Ф I is 90, the relay produces maximum torque. The vector diagram of these quantities for a directional relay is shown in Figure

3 Position of I for maximum torque γ = 90 β V Ф I I Ф V Figure 3.3 Vector diagram for directional relay The operating torque (T) can be expressed as follows: operating Ф V Ф I sin γ (3.1) Since, Ф I I and Ф V V operating V I sin γ From the vector diagram, it can be seen that + = 90 = 90 and γ + β =, γ = β operating V I sin( ) V I sin(90 ) V I sin 90 ( ) (3.2) V I cos( ) (3.3) From the phasor, the maximum torque angle is given by = 90. If the angle ( + β) is less than +90 and more than 90, the torque will be positive and overturn results in negative torque. The directional relay measures the angle between the voltage and current to identify the correct directional of current flow away from the bus. In case of fault near the bus, the voltage available on PT secondary is not enough to produce an operating torque in directional relay (electromagnet). This voltage value depends on the location of the fault on line from the relaying point. The minimum fault distance from the relay point for which the relay fails to operate is known as dead zone. Figure 3.4 shows the 43

4 characteristic of directional relay with dead zone [16]. Operating direction Blocking direction Directional relay characteristic Dead zone Figure 3.4 Directional relay characteristic Polarizing Quantity Directional relay is a two-quantity relay, and it compares the phase angle of the input voltage and current quantities. The directional overcurrent relay operates only when the magnitudes of current become higher than the set value of the threshold, and the current flows in its correct operating direction (forward direction). The torque produced in the directional overcurrent relay is maximum when = β. During fault, the power factor angle is large i.e. of the orders of 80 to 90 and it depends on the location of fault.hence, the maximum torque angle = 90 should be of the same order to achieve maximum torque in the relay during fault. The maximum torque angle can be set to 30, 60, and 90 by suitable connection of CTs and PTs in the relaying circuit. Connections of 30 offer negative torque and maloperation of the directional relaying scheme for certain types of faults. Connections of 60 produce low torque for certain types of faults. Hence, 30 and 60 connections are not widely used for directional relaying scheme. In 90 connections, the polarizing voltage is fed to phase element in such way that it produces maximum torque. In case of R B (L L) fault, the voltage across the R element is V YB and across the B element is V RY. Thus, the required maximum torque is produced by providing the polarizing voltage of healthy phase to the voltage coil of active (faulted) phase. Figures 3.5(a) and 3.5(b) show the 90 connection and the vector diagram for R B fault, respectively. 44

5 I R V RY V R V R V BR V RY V y I R V B V B I B (a) 90 V YB I Y V B I B = I R (b) V Y V YB Figure 3.5 R B fault (a) 90 connection (b) Vector diagram In case of unity power factor, the position of faulted phase current (I R ) leads the polarizing voltage (V YB ) by 90. In the event of high resistance fault, the directional overcurrent relay with 90 connections produces less torque. Hence, to achieve maximum torque, the maximum torque angle can be adjusted to any desired value by inserting a resistance or capacitance in series with the voltage coil of the directional relay. Table 3.1 shows the various combinations of voltages and current fed to the directional relays for 30, 60, and 90 connections [16]. Table 3.1 Quantities fed to phase element of directional relay Types of connections Fault involving phase R Fault involving phase Y Fault involving phase B Current Voltage Current Voltage Current Voltage 30 I R V RB I Y V YR I B V BY 60 I R I Y V RB I Y I B V YR I B I R V BY 90 I R V YB I Y V BR I B V RY Directional Ground-Fault Relays A residual current and residual voltage are fed to the directional ground-fault relay. The value of residual voltage (V R + V Y + V B ) is zero for normal operating condition as well as during phase faults. During ground fault, the residual voltage of open delta PT secondary operates the directional relay. Figure 3.6 shows the vector diagram for L G (R G) ground fault. 45

6 V R V R V Y I R V Res V B V Res = V R + V Y + V B I Res = I R + I Y + I B V B V Y Figure 3.6 Vector diagram for L G (R G) ground fault 3.2 CURRENT STATE OF THE ART In order to achieve coordination of overcurrent relays in ring-fed distribution network containing DG, So et al. [26] presented time coordinate method based on evolutionary programming. However, the prime limitation of this method is that it cannot handle many fault current redistributions at a time. After that, Jager et al. [60] suggested a coordination method for ring and radial distribution network containing DG. But the above scheme suffers with the problem of sympathy trips depending upon the location and level of fault [28]. Brito et al. [135] discussed the impact of the insertion of DG in the protection coordination using simulation software. James et al. [64] presented an analysis of several protection coordination problems such as fault detection ability, the characteristic contribution of fault current, effects of increased short circuit capacities and islanding of DGs due to integration of DGs into the electric distribution system. However, no malfunctioning of protection device has been reported due to incorporation of DGs. Thereafter, Rifat [109] has given several considerations for utility/cogeneration inter-tie protection scheme. But these considerations are mainly for distance relays. Salman et al. [124] presented an investigation on the impact of the integration of embedded synchronous generators and embedded induction generators on the settings of protective devices installed on the distribution systems. Afterwards, several case studies related to the impact of DG on protection system have been presented by different authors [55], [23], [145]. It has been observed that the conventional distribution protection schemes do not have sufficient capability to protect radial distribution network with all possible configurations and operating conditions of DG [110]. 46

7 Tong and his co-worker [45] presented a concept of FCL (Fault Current Limiter) to limit the effect of DG on the coordinated relay protection scheme in a radial system. But the proposed concept can be applicable only up to certain tolerable levels with reference to tripping time. Manjula et al. [79] proposed an inverse time admittance relay based on line admittance measurement to protect a distribution network with converter interfaced DG. However, relay operation may be restrained beyond predefined value of fault resistance. Further, the cost of distance relays is comparatively higher than overcurrent relays. However, none of these schemes completely solved the problem of miscoordination between relay in radial distribution systems in the presence of DG. In order to avoid most of the above drawbacks, an attempt has been made to demonstrate the concept of directional relay for the protection of radial distribution network containing DGs. The proposed work was supported with the developed prototype of 3-phase radial distribution system in laboratory environment along with their comparative evaluation with the results obtained using PSCAD/EMTDC software package. 3.3 THE PROPOSED DIRECTIONAL RELAYING SCHEME Figure 3.7 shows a single line diagram of the proposed directional relaying scheme. The relays nearest to the utility and DG2 (R 1 and R 4 ) are non-directional whereas the remaining relays (R 2 and R 3 ) are directional in nature as the fault current at a particular bus changes its direction (away and toward the bus). Figure 3.7 Single line diagram of the proposed directional relaying scheme with DG 47

8 The directional feature is gained by comparing the direction of current flow in the line with reference to the bus voltage. Thus, the directional overcurrent relay operates only when the current flowing through the relay is in its correct direction and more than its pickup setting. With reference to Figure 3.7, for a fault in section-1, relays R 1 and R 2 operate and disconnect the section-1 only, while the section-2 remains in healthy condition. During the implementation of the proposed scheme in the laboratory environment and also in the computer simulation, the following assumptions have been carried out. 1. In the development of laboratory prototype, authors have used class one CTs and PTs whereas for computer simulation in PSCAD/EMTDC, available Jiles-Atherton CT model has been used. 2. The effect of saturation of CT on the performance of the proposed scheme during computer simulation in PSCAD/EMTDC software package has not been considered. On the other hand, the effect of CT saturation, depending upon the magnitude of fault current, will affect the performance of relays used in developed laboratory prototype. 3. Relays used in the developed laboratory prototype are electromechanical in nature whereas in the computer simulation, modules of static relays having similar characteristic to that of electromechanical relays used in developed laboratory prototype are used. 4. In the proposed developed laboratory prototype, state electricity board supply has been used as utility. On the other hand, three-phase synchronous generators having different capacities have been used as DGs. As some system structure is required to produce realistic data, a small portion of large power distribution system has been used. Due to practical limitations, only two sections for the implementation of the proposed scheme in the laboratory have been chosen. However, the idea of the proposed directional protection scheme can be further extended to larger distribution network. 3.4 DEVELOPED LABORATORY PROTOTYPE OF THE PROPOSED DIRECTION RELAYING SCHEME Figure 3.8 shows laboratory prototype of the proposed directional relaying scheme. Synchronous generator has been used as DG which is connected at Bus-B and Bus-C whereas 48

9 the utility supply is connected at Bus-A. Radial distribution network is simulated using two impedances namely Z 1 and Z 2. Star connected load bank is used as 3-phase load at Bus-B and Bus-C. Two 3-phase tap loads (1 and 2) are taken from the middle of the section-1 and section-2. Faults are simulated using four toggle switches namely S1, S2, S3 and S4. Variable rheostat Rh1 is used as a fault current limiter for practical purposes. All the four relays are connected in secondary side of Current Transformers (CTs). CT shorting switch is used to check the back-up protection of a particular section of radial distribution network. Circuit breakers are simulated by contactors. Different loads of the distribution network are protected by Miniature Circuit Breakers (MCBs) M1, M2, M3 and M4 respectively. In order to protect DG against short-circuit, another MCB M5 for DG2 and M6 for DG1 are used. Power circuit of DG1 is same as DG2. Moreover, relays, CTs and impedances are also connected in other Figure 3.8 Power circuit of laboratory prototype of the proposed directional relaying scheme 49

10 two phases. However, due to complexity, it has not shown in Figure 3.8. Detailed specifications of the different components used in the said laboratory prototype are given in Appendix-B. Figure 3.9 shows control circuit of the said laboratory prototype. Looking to Figure 3.8 and Figure 3.9 together, by pressing the spring loaded push button of section-1 (PB1), contactor (C1) energises. The hold on path is being provided by auxiliary contact of contactor C1-2 and hence, C1 remains in energised condition. Thus, contact C1-1 in Figure 3.8 remains in close condition until contactor C1 will de-energise. The contactor (C1) is de- energised using stop push button of section-1 (PB2). The magnitude of fault current is very high and depends on the location of fault, capacity of DG and fault resistance. For a fault in section-1 (by closing switch S1), the fault current referred to secondary of CT1 exceeds pickup setting of the relay R 1, and hence, relay R 1 operates. Closing of relay contact R 1-1 energises auxiliary relay A1 in the control circuit (Figure 3.9) and finally, the contact of auxiliary relay (A1-1) opens out. Opening of A1-1 deenergises the contactor (C1), and hence, disconnecting the section-1 from the utility supply side due to opening of the contact C1-1 (Figure 3.8). When the relay contact R 1-1 opens out, coil of auxiliary relay A1 remains energized due to hold on path provided by contact A1-2 (Figure 3.9). The coil of auxiliary relay A1 de-energises when reset push button PB3 is pressed. For the same fault in section-1, due to contribution in the fault current from DGs, current in secondary of CT2 also exceeds pickup setting of the directional relay R 2. Therefore, relay R 2 operates which closes the relay contact R 2-1. This energises auxiliary relay A2 in the control circuit (Figure 3.9) and hence, auxiliary relay contact A2-1 opens out. Opening of A2-1 de-energises the contactor (C2), and hence, disconnecting the section-1 from DG side due to opening of the contact C2-1 (Figure 3.8). When relay contact R 2-1 opens out, coil of auxiliary relay A2 remains energized due to hold on path provided by its own contact A2-2 (Figure 3.9). The coil of auxiliary relay A2 de-energises when reset push button PB6 is pressed. Similarly, the same procedure is followed for a fault in section-2 (Figure 3.8). Figure 3.10 shows pictorial view of the developed laboratory prototype for three phase radial distribution network having two sections. 50

11 Figure 3.9 Control circuit of laboratory prototype of the proposed directional relaying Scheme Figure 3.10 Developed laboratory prototype of three phase radial distribution network of the proposed directional relaying scheme 51

12 3.5 SOFTWARE SELECTION FOR MODELING OF THE SYSTEM The software selected for this purpose is PSCAD (Power System Computer Aided Design) which uses EMTDC (Electromagnetic Transient in DC System) [106]. This software has the capability of performing interpolation between minimum time steps. It uses trapezoidal methods for solving numerical integrations and differential equations. It manifests the continuous oscillation of the node voltage (branch current) with changing direction in every time step, which is not a representative of any electrical behavior of the network. It enables the user to schematically construct a circuit, run a simulation, analyse the results, and manage the data in a completely integrated, graphical environment. Online plotting functions, controls, and meters are also included so that the user can alter system parameters during a simulation run and view the results directly. It comes with a library of pre-programmed and tested models, ranging from simple passive elements and control functions to more complex models, such as electric machines, FACTS (flexible AC transmission systems) devices, transmission lines, relays, cables and many power system devices. If a particular model does not exist, it provides the flexibility of building custom models, either by assembling those graphically using existing models, or by utilizing an intuitively designed design editor. 3.6 MODELING AND SIMULATION OF RADIAL DISTRIBUTION SYSTEMS IN PRESENCE OF DG USING PSCAD AND LABORATORY PROTOTYPE A part of the radial distribution system, as shown in Figure 3.7, has been used to access the problems associated with radial system containing DGs and also to validate the proposed scheme. Figure 3.11 Simulation model of radial feeder with DG 52

13 The distribution line parameters and the generating station details are given in Appendix-C. Test data for verifying the proposed scheme have been generated by modeling the complete system of Figure 3.7 using PSCAD/EMTDC software package [106]. Figure 3.11 shows simulated model of radial feeder containing DG with two bidirectional relays (R 1 and R 4 ) and two directional relays (R 2 and R 3 ). The distribution feeder is represented using the Bergeron line model. For modeling the whole system, some of the main library components in PSCAD such as utility source, DG and circuit breakers etc. have been used. They are designed according to the collected data and specifications. Further, some self-created components such as directional relays have been developed in PSCAD. Further, the updating of system parameters is achieved using Multi-Run block available in PSCAD to generate multiple simulation cases in single shot. The relays, as shown in Figure 3.7, are located at each end of the distribution feeder. PS and TMS of each relay are calculated using IEC standard relay characteristics equation. Plug setting of each relay is based on full load current of the feeder during normal/prefault condition. Time Multiplier Setting of relay R 1, R 2, R 3 and R 4 are calculated based on fault calculations. We have achieved 0.2, 0.1, 0.1 and 0.2 TMS for relay R 1, R 2, R 3 and R 4, respectively. However, in order to avoid practical limitations, we have set 0.7, 0.6, 0.6 and 0.7 TMS for relay R 1, R 2, R 3 and R 4, respectively [18]. The performance of the proposed scheme has been evaluated for various types of faults in each section at different fault locations. 3.7 RESULT OBTAINED FROM EXPERIMENTAL PROTOTYPE AS WELL AS FROM PSCAD/ EMTDC SOFTWARE PACKAGE Performance of the Proposed Scheme during Varying Load Condition The settings of directional and non-directional overcurrent relays should be done according to the maximum balance and unbalance loading conditions. If this factor is not considered than there is a possibility of mal-operation of relays in normal conditions. Therefore, it is necessary to investigate the effect of varying load conditions on the performance of the protective devices. Table 3.2 shows the performance of the proposed scheme during varying unbalance load conditions. It has been observed from the Table 3.2 that during the said situations no relay operation has been observed. 53

14 3.7.2 Faults within Section Various types of faults have been simulated in two different sections (1 and 2) at different fault locations using fault selector switches (S1, S2, S3 and S4). Table 3.3 to Table 3.7 show results obtained from experimental prototype as well as from PSCAD/EMTDC software package for various types of faults at different fault locations in two different sections with zero fault resistance. In all the tables, F1 and F2 indicate close-in fault in section-1 and section-2, respectively, whereas F1 and F2 indicate remote end fault in section-1 and section-2, respectively. It has been observed from Table 3.3 to Table 3.7 that for a close-in and remote end fault in the respective section (section-1 and section-2), the respective relays (R 1 & R 2 for section-1 and R 3 & R 4 for section-2) operate and disconnect the faulty section. The maloperation of bidirectional relays due to incorporation of DG has been eliminated by directional relays. The operation of relay sequence is perfectly matched with the selectivity of the protection system. It has been observed that for all types of close-in fault in section-1, the operating time (T op ) of local relay (R 1 ) is lower than the remote-end relay (R 2 ) as contribution in fault current from utility side is significant, hence R 1 operates first compare to R 2. On the other hand, for all types of remote end fault in section-1, the remote end relay (R 2 ) operates first than the local relay (R 1 ) as the contribution in fault current from both DGs are significant than the utility. Moreover, for all types of close-in and remote-end faults in section-2, the local relay (R 3 ) operates first than remote-end relay (R 4 ) as contribution in fault current from the utility and DG1 is significant than DG2. It has been observed from Table 3.3 to Table 3.7 that the time of operation of all the relays using PSCAD/ EMTDC software package is almost equal to the time of operation of relays obtained using laboratory prototype High Resistance Fault When an overhead distribution phase conductor breaks and falls on a high impedance surface or trees, high impedance fault occurs [14]. The conventional overcurrent relays at the radial distribution network may not be able to detect this type of fault and hence, relay does not operate. Table 3.8 and Table 3.9 show results obtained for single line to ground (A-g) fault at different locations in all the two sections with a fault resistance of 10 Ω and 18 Ω, respectively. Figure 3.12 shows directional relay responses, which has been obtained from PSCAD simulation, for a single line- to-ground fault at F2 in section-2 with R F = 10 Ω. 54

15 Figure 3.12 Response of directional relay for single line-to-ground fault (A-G) in section-2 with R F =10 Ω. It has been observed from Table 3.8 and Table 3.9 that the time of operation of all the relays increases as fault resistance increases. The proposed scheme gives satisfactory result in the said condition. However, the operating time of all the relays is higher with reference to low resistance fault. In certain conditions, it has been observed that the relay may not be able to pickup in case of a fault with a very high value of fault resistance. The best available solution for this situation is the usage of digital relays Backup Protection Backup protection is extremely important for any protection system if primary protection system fails to clear the fault within its own zone of protection. Backup protection feature is simulated for each section and the sample results have been shown in Table 3.8 and Table RESULT OBTAINED FROM PSCAD/ EMTDC SOFTWARE PACKAGE FOR ACTUAL TMS Time Multiplier Setting of relay R 1, R 2, R 3 and R 4 are calculated based on fault calculations. We have achieved 0.2, 0.1, 0.1 and 0.2 TMS for relay R 1, R 2, R 3 and R 4, respectively. Table 3.10 to Table 3.12 show results obtained from PSCAD/EMTDC software package for single line-to-ground fault at different fault locations in two different sections 55

16 with a fault resistance of 0 Ω, 10 Ω and 18 Ω, respectively. The performance of the proposed scheme has been evaluated for various types of faults in each section at different fault locations. It has been observed from Table 3.10 to Table 3.12 that the time of operation of all the relays decreases as TMS of all relays are less compare to results obtained from laboratory prototype with higher TMS. Moreover, the proposed scheme also provides back-up protection if primary protection system fails to operate for a fault within its own zone. Backup protection feature is simulated for each section and the sample results have been shown in Table 3.10 to Table

17 Table 3.2 Performance of the proposed scheme during variation in loading conditions Table 3.3 Result obtained for A-g faults in different sections with R F = 0 Ω 57

18 Table 3.4 Result obtained for A-B faults in different sections with R F = 0 Ω I Utility I DG1 I DG2 I Fault Plug Setting (PS) of R 1 =75%, R 2 =50%,R 3 =50% and R 4 =75% of I R (Relay rated current), Time Multiplier Setting (TMS) of R 1 =0.7, R 2 =0.6, R 3 =0.6 and R 4 =0.7,System Voltage (L-L) = 300 V, Frequency = 49.69Hz, R F = 0 ohm Current (A) Time of Operation (T op (s)) Using Laboratory Time of Operation (T op (s)) Using PSCAD Fault Prototype Location R 1 R 2 R 3 R 4 R 1 R 2 R 3 R 4 Prototype PSCAD Prototype PSCAD Prototype PSCAD Prototype PSCAD Pre-fault condition NO NO NO NO NO NO NO NO F F F F NO: No Operation 58

19 Table 3.5 Result obtained for A-B-g faults in different sections with R F = 0 Ω Plug Setting (PS) of R 1 =75%, R 2 =50%,R 3 =50% and R 4 =75% of I R (Relay rated current), Time Multiplier Setting (TMS) of R 1 =0.7, R 2 =0.6, R 3 =0.6 and R 4 =0.7,System Voltage (L-L) = 300 V, Frequency = 49.69Hz, R F = 0 ohm Time of Operation Time of Operation (T Current (A) op (s)) (T op (s)) Using Laboratory Using PSCAD Fault Prototype Location I Utility I DG1 I I Fault DG2 R 1 R 2 R 3 R 4 R 1 R 2 R 3 R 4 Prototype PSCAD Prototype PSCAD Prototype PSCAD Prototype PSCAD Pre-fault condition NO NO NO NO NO NO NO NO F F F F NO: No Operation 59

20 Table 3.6 Result obtained for A-B-C faults in different sections with R F = 0 Ω Plug Setting (PS) of R 1 =75%, R 2 =50%,R 3 =50% and R 4 =75% of I R (Relay rated current), Time Multiplier Setting (TMS) of R 1 =0.7, R 2 =0.6, R 3 =0.6 and R 4 =0.7,System Voltage (L-L) = 300 V, Frequency = 49.69Hz, R F = 0 ohm Current (A) Time of Operation Time of Operation (T (T op (s)) op (s)) Using PSCAD Using Laboratory Fault Prototype Location I Utility I DG1 I DG2 I Fault R 1 R 2 R 3 R 4 R 1 R 2 R 3 R 4 Prototype PSCAD Prototype PSCAD Prototype PSCAD Prototype PSCAD Pre-fault condition NO NO NO NO NO NO NO NO F F F F NO: No Operation 60

21 Table 3.7 Result obtained for A-B-C-g faults in different sections with R F = 0 Ω Plug Setting (PS) of R 1 =75%, R 2 =50%,R 3 =50% and R 4 =75% of I R (Relay rated current), Time Multiplier Setting (TMS) of R 1 =0.7, R 2 =0.6, R 3 =0.6 and R 4 =0.7,System Voltage (L-L) = 300 V, Frequency = 49.69Hz, R F = 0 ohm Current (A) Time of Operation Time of Operation (T (T op (s)) op (s)) Using PSCAD Using Laboratory Fault Prototype Location I Utility I DG1 I DG2 I Fault R 1 R 2 R 3 R 4 R 1 R 2 R 3 R 4 Prototype PSCAD Prototype PSCAD Prototype PSCAD Prototype PSCAD Pre-fault condition NO NO NO NO NO NO NO NO F F F F NO: No Operation 61

22 Table 3.8 Result obtained for A-g faults in different sections with R F =10 Ω Plug Setting (PS) of R 1 =75%, R 2 =50%,R 3 =50% and R 4 =75% of I R (Relay rated current), Time Multiplier Setting (TMS) of R 1 =0.7, R 2 =0.6, R 3 =0.6 and R 4 =0.7,System Voltage (L-L) = 300 V, Frequency = 49.69Hz, R F =10 ohm Current (A) Time of Operation (T op (s)) Using Laboratory Time of Operation (T op (s)) Using PSCAD Fault Prototype Location Pre-fault condition I Utility I DG1 I DG2 I Fault R 1 R 2 R 3 R 4 R 1 R 2 R 3 R 4 Prototype PSCAD Prototype PSCAD Prototype PSCAD Prototype PSCAD NO NO NO NO NO NO NO NO F F F F Backup Protection For First Section (F1 ) Backup Protection For Second Section (F2 ) NO: No Operation If Relay R 2 Fails If Relay R 3 Fails If Relay R 2 Fails If Relay R 3 Fails

23 Table 3.9 Result obtained for A-g faults in different sections with R F = 18 Ω 63

24 Table 3.10 Simulation result obtained for A-g faults in different sections with R F = 0 Ω Plug Setting (PS) of R 1 =75%, R 2 =50%,R 3 =50% and R 4 =75% of I R (Relay rated current) Time Multiplier Setting (TMS) of R 1 =0.2, R 2 =0.1, R 3 =0.1 and R 4 =0.2 System Voltage (L-L) = 300 V, Frequency = 50 Hz, R F =0 Ω Fault Location Current (A) Time of Operation (T op (s)) Using Pscad I Utility I DG1 I DG2 I Fault R 1 R 2 R 3 R 4 Pre-fault condition NO NO NO NO F F F F Backup Protection For First Section (F1 ) Backup Protection For Second Section (F2 ) NO: No Operation If Relay R 2 Fails If Relay R 3 Fails

25 Table 3.11 Simulation result obtained for A-g faults in different sections with R F = 10 Ω Plug Setting (PS) of R 1 =75%, R 2 =50%,R 3 =50% and R 4 =75% of I R (Relay rated current) Time Multiplier Setting (TMS) of R 1 =0.2, R 2 =0.1, R 3 =0.1 and R 4 =0.2 System Voltage (L-L) = 300 V, Frequency = 50 Hz, R F =10 Ω Fault Location Pre-fault condition Current (A) Time of Operation (T op (s)) Using Pscad I Utility I DG1 I DG2 I Fault R 1 R 2 R 3 R NO NO NO NO F F F F Backup Protection For First Section (F1 ) Backup Protection For Second Section (F2 ) NO: No Operation If Relay R 2 Fails If Relay R 3 Fails

26 Table 3.12 Simulation result obtained for A-g faults in different sections with R F = 18 Ω Plug Setting (PS) of R 1 =75%, R 2 =50%,R 3 =50% and R 4 =75% of I R (Relay rated current) Time Multiplier Setting (TMS) of R 1 =0.2, R 2 =0.1, R 3 =0.1 and R 4 =0.2 System Voltage (L-L) = 300 V, Frequency = 50 Hz, R F =18 Ω Fault Location Pre-fault condition Current (A) Time of Operation (T op (s)) Using Pscad I Utility I DG1 I DG2 I Fault R 1 R 2 R 3 R NO NO NO NO F F F F Backup Protection For First Section (F1 ) Backup Protection For Second Section (F2 ) NO: No Operation If Relay R 2 Fails If Relay R 3 Fails

27 3.9 MODELING AND SIMULATION OF A LARGE 11 KV RADIAL DISTRIBUTION SYSTEMS IN THE PRESENCE OF DG USING PSCAD System Description of a Large 11 kv Radial Distribution Systems in the Presence of DG A part of the Indian 11 kv radial distribution system, as shown in Figure 3.7 and Figure 3.8, has been used to access the problems associated with radial system containing DGs and also to validate the proposed scheme. The distribution line parameters and the generating station details are given in Appendix-D. Test data for verifying the proposed scheme have been generated by modeling the complete system of Figure 3.7 using the PSCAD/EMTDC software package [106]. The performance of the proposed scheme has been evaluated for various types of faults in each section at different fault locations. Relay responses for some special cases such as high resistance fault and backup protection were also investigated Modeling of Bidirectional Relays in PSCAD Figure 3.13 shows simulated model of radial feeder containing DG with the conventional (bidirectional) relays. Figure 3.14 shows the tripping logic of bidirectional relays. With reference to Figure 3.14, when a current through a bidirectional relay (51) exceeds the pick-up setting of relay (details are given in Appendix-D), it generates a constant signal (logic 1) through Hysteresis-buffer. The output of the respective phase relay is given to OR gate. Depending upon the output (logic 1/logic 0) of any phase relay, OR gate generates a trip signal which will be further given to the respective breakers of the respective sections (B1, B2, B3). In order to check back up protection of each section, two state (0-off, 1-on) selector switch is used which bypass the relay of a particular section of the respective phase. Figure 3.13 Simulated model of bidirectional relay 67

28 Figure 3.14 Tripping logic of bidirectional relay Modeling of Directional Relays in PSCAD Figure 3.15 shows simulated model of radial feeder containing DG at Bus-2, Bus-3 and Bus-4 with two bidirectional relays (R 1 and R 6 ) and four directional relays (R 2, R 3, R 4 and R 5 ). Figure 3.16 shows the tripping logic of directional relays. In this logic, bidirectional relays are made directional through a logic circuit. Generally, 90 connection scheme is used in two phase overcurrent and one earth fault protection scheme. The directional feature is accomplished by comparing the phase angle between the current of the faulted phase and line voltage of healthy phase in phase comparator block. For example, for fault in phase-a in section-1, phase angle (P2A) of current (IT2) in phase A is compared with phase angle (E2B) of line voltage (Eb2) of other two lines (Figure 3.15 & 3.16).The FFT (Fast Fourier Transformations) block has been used in order to obtain phase angle between the voltage and the current quantity of the respective phase. Figure 3.15 Simulated model of directional relay 68

29 Figure 3.16 Tripping logic and phase comparator block of directional relay Line currents of each section are fed to the respective relay of the respective phase through CTs. As shown in Figure 3.16, in order to determine the region of operation of directional relay, the phase difference between current and voltage which is shifted by +90 is fed to the trigonometric COS function. If this phase difference is within the forward limit than the output of AND gate of the respective phase (T2A) is made positive (high) which is used to activate the two-state switch. With this situation of two-state switch, it allows the CT secondary current to pass through the respective relay of that phase (phase A) otherwise twostate switch remains inoperative. If fault current is in forward direction and line current fed to the relay from respective CT through two-state switch exceeds the pick-up setting than the respective directional relay generates a spike which will be made constant by Hysteresis- Buffer. Thereafter, the output of each phase relay is given to OR gate. Depending upon the output (logic 1) of the respective phase relay, OR gate generates a final tripping signal which will be given to the respective breaker of the respective section (B2, B3, B4, B5). To check the backup protection in a particular section, another ON- OFF switch (R2A) is used in phase comparator block. Trip characteristic constants of each relay are set in such a way that relay gives normal inverse characteristic. PS and TMS of each relay is set such that relay R 3 gives back up to relay R 5 and relay R 1 gives back up to relay R 3 for downstream faults. Similarly relay R 4 gives back up to relay R 2 and relay R 6 gives back up to relay R 4 for upstream faults. 69

30 3.10 RESULTS AND DISCUSSION Bidirectional Relay Table 3.13 shows simulation results obtained in terms of fault currents and Time of Operation (T op ) of three bidirectional relays for faults at different locations in three different sections with zero fault resistance. The radial feeder containing DG is protected with three bidirectional over current relays namely R 1, R 2 and R 3 as shown in Figure The fault locations F1, F3 and F5 indicate close-in fault whereas F2, F4 and F6 indicate remote end fault in section-1, section-2 and section-3, respectively. Table 3.13 Simulation results of bidirectional relay for L-G faults in different sections with R F = 0 Ω Plug Setting (PS) of R 1 =100% of I R, R 2 =75% of I R, R 3 =50% of I R (Relay Rated Current = 1 A) Time Multiplier Setting (TMS) of R 1 =0.15, R 2 =0.125, R 3 =0.1 Time of Operation Fault R F = 0 Ω (Top (s)) Location I U (A) I DG (A) I F (A) Pre-fault condition A B C A B C A B C R 1 R 2 R NO NO NO F F F F F F Backup Protection For Third Section (F5) NO: No Operation If Relay R 3 fails 70

31 It has been observed from Table 3.13 that for a close-in and remote end fault in section-1 (F1 and F2), relay R 1 operates first as the contribution of fault current from the utility side (I U ) is significant (strong source). On the other hand, relay R 3 operates after some time delay as the contribution of fault current from DG (weak source) is comparatively lower than the utility. With the present structure of radial distribution network along with DG, relay R 1 and relay R 2 has to trip for a fault in section-1. But, the breaker in section-3 trips unnecessary due to operation of relay R 3. This mal-operation occurs due to the presence of DG on the other side, which is against the selectivity of the protection system. Correspondingly, for a close in and remote end fault in section-2 (F3 & F4) relay R 2 and relay R 3 operate whereas for a close in and remote end fault in section-3 (F5 & F6), relay R 3 and local protection of DG (Fuse/MCB) operate. It is to be noted that as the relay is located near the DG, the contribution in fault current from DG is higher. This would not create a problem in maintaining selectivity as one can easily disconnect DG when its penetration is low. On the other hand, when penetration of DG is high and more than one DG is connected at different location, the conventional relay is not in a position to maintain the proper selectivity. This problem is rectified using the proposed scheme which works perfectly even during different DG capacities and also during more than one DG connected at various locations Directional Relay Table 3.14 to Table 3.17 show the simulations results for L-G, L-L-G, L-L-L-G and L-L faults at different fault locations in three different sections with zero fault resistance. The radial feeder containing DG is protected with four directional over current relays R 2, R 3, R 4, R 5 and two bidirectional relays R 1 and R 6 as shown in Figure It has been observed from Table 3.14 to Table 3.17 that for a fault in section-1, section-2 and section-3 at any locations relay R 1 & relay R 2, relay R 3 & relay R 4, relay R 5 & relay R 6 operate and disconnect the faulty section, respectively. The maloperation of bidirectional relays due to incorporation of DG has been eliminated by the proposed scheme. The operation of relay sequence is perfectly matched with the selectivity of the protection system. Further, it has been observed from Table 3.14 to Table 3.17 that for all close-in and remote-end faults in all the three sections, the local end relay (R 1, R 3 and R 5 ) operates first than the remote-end relay (R 2, R 4 and R 6 ) as the contribution of fault current from the utility (I U ) is significant (strong source). Figure 3.17 shows directional relay responses, which has been obtained from PSCAD simulation, for a single line- to-ground fault at F4 in section-2 with R F = 0 Ω. 71

32 Figure 3.17 Response of directional relay for single line-to-ground fault (A-G) in section-2 with R F = 0 Ω Backup Protection Backup protection is extremely important for any protection system if primary protection system fails to clear the fault within its own zone of protection. Backup protection feature is simulated for section-3 and the sample result has been shown in Table High Resistance Fault When an overhead distribution phase conductor breaks and falls on a high impedance surface or trees, high impedance fault occurs [14]. The conventional overcurrent relays at the radial distribution network may not be able to detect this type of fault and hence, relay does not operate. To analyze this condition, a case study has been set up and a single line-toground fault with a fault resistance equal to 15 has been simulated. However, in practice, this value may exceed 30 Ω. Table 3.18 shows the simulation results for L-G faults at different locations in three different sections with a fault resistance of 15 Ω. It has been observed from Table 3.18 that for a fault in any section, the primary protection relay of that section operates first and clears the fault within its own zone. Though, a slower relay performance is achieved, the proposed scheme operates successfully. 72

33 Table 3.14 Simulation results of directional relay for L-G faults in different sections with R F = 0 Ω Plug Setting (PS) of R 1 =50% of I R, R 3 =75% of I R, R 5 =100% of I R, R 2 =100% of I R, R 4 =75% of I R, R 6 =50% of I R (Relay Rated Current = 1 A) Time Multiplier Setting (TMS) of R 1 =0.1, R 3 =0.125, R 5 =0.15, R 2 =0.15, R 4 =0.125, R 6 =0.1 Fault R F = 0 Ω Time of Operation (Top (s)) Location I U (A) I DG 2 (A) I DG3 (A) I DG4 (A) I F (A) A B C A B C A B C A B C A B C R 1 R 2 R 3 R 4 R 5 R 6 Pre - fault condition NO NO NO NO NO NO F F F F F Back up protection If R 5 fails F NO: No Operation 73

34 Table 3.15 Simulation results using directional relay for L-L-G faults in different sections with R F = 0 Ω Plug Setting (PS) of R 1 =50% of I R, R 3 =75% of I R, R 5 =100% of I R, R 2 =100% of I R, R 4 =75% of I R, R 6 =50% of I R (Relay Rated Current = 1 A) Time Multiplier Setting (TMS) of R 1 =0.1, R 3 =0.125, R 5 =0.15, R 2 =0.15, R 4 =0.125, R 6 =0.1 Fault R F = 0 Ω Time of Operation (Top (s)) Location I U (A) I DG2 (A) I DG3 (A) I DG4 (A) I F (A) A B C A B C A B C A B C A B C R 1 R 2 R 3 R 4 R 5 R 6 Pre - fault NO NO NO NO NO NO condition F F F F F F NO: No Operation 74

35 Table 3.16 Simulation results using directional relay for L-L-L-G faults in different sections with R F = 0 Ω Plug Setting (PS) of R 1 =50% of I R, R 3 =75% of I R, R 5 =100% of I R, R 2 =100% of I R, R 4 =75% of I R, R 6 =50% of I R (Relay Rated Current = 1 A) Time Multiplier Setting (TMS) of R 1 =0.1, R 3 =0.125, R 5 =0.15, R 2 =0.15, R 4 =0.125, R 6 =0.1 Fault R F = 0 Ω Time of Operation (Top (s)) Location I U (A) I DG2 (A) I DG3 (A) I DG4 (A) I F (A) A B C A B C A B C A B C R 1 R 2 R 3 R 4 R 5 R 6 Pre - fault NO NO NO NO NO NO condition F F F F F F NO: No Operation 75

36 Table 3.17 Simulation results using directional relay for L-L faults in different sections with R F = 0 Ω Plug Setting (PS) of R 1 =50% of I R, R 3 =75% of I R, R 5 =100% of I R, R 2 =100% of I R, R 4 =75% of I R, R 6 =50% of I R (Relay Rated Current = 1 A) Time Multiplier Setting (TMS) of R 1 =0.1, R 3 =0.125, R 5 =0.15, R 2 =0.15, R 4 =0.125, R 6 =0.1 Fault R F = 0 Ω Time of Operation (Top (s)) Location I U (A) I DG2 (A) I DG3 (A) I DG4 (A) I F (A) A B C A B C A B C A B C A B C R 1 R 2 R 3 R 4 R 5 R 6 Pre - fault NO NO NO NO NO NO condition F F F F F F NO: No Operation 76

37 Table 3.18 Simulation results using directional relay for L-G faults in different sections with R F =15 Ω Plug Setting (PS) of R 1 =50% of I R, R 3 =75% of I R, R 5 =100% of I R, R 2 =100% of I R, R 4 =75% of I R, R 6 =50% of I R (Relay Rated Current = 1 A) Time Multiplier Setting (TMS) of R 1 =0.1, R 3 =0.125, R 5 =0.15, R 2 =0.15, R 4 =0.125, R 6 =0.1 Fault R F =15 Ω Time of Operation (Top (s)) Location I U (A) I DG2 (A) I DG3 (A) I DG4 (A) I F (A) A B C A B C A B C A B C A B C R 1 R 2 R 3 R 4 R 5 R 6 Pre - fault NO NO NO NO NO NO condition F F F F F F NO: No Operation 77

38 3.11 EFFECT OF CHANGE IN CAPACITY OF DG In order to analyze the effect of penetration of DG on the proposed scheme, different capacities of DG in terms of total load on the radial feeder have been considered. Hence, 21% capacity of DG indicates that out of total load on the feeder, DG supplies 21% whereas 79% of total load is supplied by the utility. The simulation has been carried out for four different DG capacities viz., 21%, 33%, 50% and 59% of the total load on the feeder. Figure 3.18 and Figure 3.19 show the simulation results in the form of time of operation of relay R 1 and relay R 2 for the said four different DG capacities for L-G fault (minimum fault current) and L-L-L-G fault (maximum fault current) at location F1 & F2. The simulation results in the form of time of operation of relay R 3 and relay R 4 for the said four different DG capacities for L-G fault (minimum fault current) and L-L-L-G fault (maximum fault current) at location F3 & F4 is shown in Figure 3.20 and Figure 3.21, respectively. Figure 3.22 and Figure 3.23 show the simulation results in the form of time of operation of relay R 5 and relay R 6 for the said four different DG capacities for L-G fault (minimum fault current) and L-L-L-G fault (maximum fault current) at location F5 & F6. It has been observed from Figure 3.18 to Figure 3.23 that, as the penetration of DG increases, time of operation of relays R 1, R 3 and R 5 increases whereas time of operation of relays R 2, R 4 and R 6 reduces. The proposed scheme gives satisfactory result in the said condition. Figure 3.18 Effect of change in capacity of DG on time of operation of relay R 1 and relay R 2 for L-G fault at F1 & F2 78