Synchrophasor-Based Backup Differential Protection Design of Transmission Lines

Transcription

1 Synchrophasor-Based Backup Differential Protection Design of Transmission Lines Prepared by Yaojie Cai Keke Jin Grace Ngantian Jiajing Wang Final Report submitted in partial satisfaction of the requirements for the degree of Bachelor of Science in Electrical and Computer Engineering in Faculty of Engineering of the University of Manitoba Faculty Supervisor: Dr. Athula Rajapakse Spring 2014 Copyright by Yaojie Cai, Keke Jin, Grace Ngantian, and Jiajing Wang, 2014

2 Abstract When power system disturbances occur, they can lead to large scale power outages if these disturbances are ignored and no action is taken to clear them. In the past few years, many blackouts have happened as a result of poor fault detection methods implemented over a wide area. With the objective of improving these methods, this report describes the design and development of a synchrophasor-based backup differential protection system using transmission line monitoring to detect faults. For experimental purposes, the power system portion of the proposed synchrophasor network is simulated using a real-time digital simulator (RTDS). The RTDS generates analog current and voltage outputs corresponding to the designed power system in real time, and they are used as inputs to phasor measurement units (PMUs). These units take phasor measurements of the simulated transmission line at different locations synchronously through the use of a Global Positioning System (GPS) clock thus, the measurements are referred to as synchrophasors. The synchrophasor data is then used to detect faults by applying a detection algorithm to the data, which will also pinpoint the location of the fault along the line. If a fault is detected, a backup signal produced by a microcontroller will be sent to trip the breakers in order to isolate the fault. To test the design, the RTDS was used to simulate various types of faults. The test results verify that the designed synchrophasor-based protection system is highly accurate and reliable. Therefore, it is applicable to modern power systems. ii

3 Contributions The advancement of Synchrophasor-based Protection will contribute to wide area monitoring of transmission line. This project can be divided into four parts: 1. Model building and Hardware Configuration 2. Developing Fault Detection Algorithm and Window Application 3. Designing and Implementation Fault Locator Algorithm 4. Designing the Backup Protection Implementation Dr. A. Rajapakse is our faculty advisor. The project was conceived by him. He provided valuable experience, knowledge and guidance to us throughout the project. Rubens Eduardo Almiron Bonnin, Naushath Monhamed Haleem, and Dinesh Gurusinghe provided us with help and valuable advice regarding academic theory. Grace was responsible for implementing a power system model in RTDS simulator with analog outputs suitable to interface with PMUs. Yaojie was responsible for implementing a differential protection algorithm for fault detection and develop a window application for displaying fault-condition results. Keke Jin was responsible for testing and implementing Phasor-based Method to pinpoint the location of fault in transmission line. Jiajing was responsible for designing a method for utilizing microcontroller to send alarm signal for indicating fault through an IP network. iii

4 Acknowledgments It is with immense gratitude that we acknowledge the support and help of our Professor Athula Rajapakse. It was an honour to receive advice from such a knowledgeable person. We must also thank Rubens Eduardo Almiron Bonnin, Naushath Monhamed Haleem, Dinesh Gurusinghe and other academic, technical staff in the power research group. Without their guidance and persistent help this project would not have been possible. We would like to thank our examining committee for spending the time to review our thesis. Yaojie Cai Keke Jin Grace Ngantian Jiajing Wang iv

5 Nomenclature Abbreviations RTDS: Real Time Digital Simulator PMU: Phasor Measurement Unit GPS: Global Positioning System PDC: Phasor Data Concentrator SEL: Schweitzer Engineering Laboratories CT: Current Transformer PT: Potential Transformer CVT: Capacitive Voltage Transformer GTAO: Giga Transceiver Analog Output GPC: Giga-Processor Card TCP/IP: Transmission Control Protocol / Internet Protocol UDP: User Datagram Protocol LED: Light-Emitting Diode Symbols : Total current of Phase A from vector summation of both ends of Transmission line. : Total current of Phase B from vector summation of both ends of Transmission line. : Total current of Phase C from vector summation of both ends of Transmission line. : Phase A current from sending end of Transmission line. : Phase B current from sending end of Transmission line. : Phase C current from sending end of Transmission line. : Phase A current from receiving end of Transmission line. : Phase B current from receiving end of Transmission line. : Phase C current from receiving end of Transmission line. : Total current between Phase A and C from vector summation of both ends of Transmission line. : Total current between Phase B and C from vector summation of both ends of Transmission line. : Total current between Phase C and A from vector summation of both ends of Transmission line. A complex number which has 1 amplitude and phase is 120 degree. : Total zero sequence current from vector summation of both ends of Transmission line. v

6 : Total positive sequence current from vector summation of both ends of Transmission line. : Total negative sequence current from vector summation of both ends of Transmission line. : The angle of total zero sequence current. : The angle of total negative sequence current. The magnitude of total current between Phase B and C. : The maximum number among A, B and C. K: Restraint coefficient a: Assumed magnitude of phase current from sending end of transmission line. b: Assumed phase difference of phase current between sending end and receiving end of transmission line. : Estimated distance to the fault location. dexact :Exact distance to the fault location. : Total line length of transmission line. X: Distance to fault location from receiving end of transmission line. Voltage at fault location. : post fault voltages at sending : post fault voltages at receiving end : Post fault currents at sending ends Post current receiving ends : Characteristic impedance; γ: Propagation constant. Coefficients of differential equation of voltage in Long Line Transmission Model. vi

7 List of Figures Figure 1.1: Main Setup Figure 2.1: Preliminary Power System Model in RSCAD Figure 2.2: Secondary Power System Model in RSCAD Figure 2.3: Secondary Model in PSS/E Figure 2.4: Current and Voltage Inputs of the SEL-421 Figure 2.5: GTAO Card Figure 2.6: RTDS/PMU Interface Figure 3.1: False Warning Window under Phase A and B Line to Ground Fault Figure 3.2: Current vs Package Index Plot under Phase A and B Line to Ground Fault Figure 3.3: Detecting Fault Time for Current Magnitude Threshold and Phase Selection Method Figure 3.4: Detecting Fault Time for Current Scalar Summation Method Figure 4.1: Transmission Long Line Model Figure 4.2: Unsaturation Case of Estimated Current (IAs) vs Actual Current (IBRKA2) Figure 4.3: Post Fault Voltage Curve Figure 4.4: Post Fault Current Curve Figure 4.5: Real Parts of Voltage Phase A on the Sending End of Transmission Line Figure 4.6: Imaginary Parts of Voltage Phase A on the Sending End of Transmission Line Figure 4.7: Flowchart of Matlab program for calculating the Fault location Figure 4.8 3: Phase Fault -Measured Error percentage vs Actual Fault location Figure 4.9: LG Fault -Measured Error percentage vs Actual Fault location Figure 4.10: LLG Fault -Measured Error percentage vs Actual Fault location vii

8 Figure 4.11: Various Fault Impendence -Measured Error percentage vs Actual Fault location Figure 5.1: SEL-421 Functional Overview Figure 5.2: Original Breaker Logic Circuit Figure 5.3: A Complete Program for the Microcontroller Interface method Figure 5.4: Program for the LED Demonstration Figure 5.5: Program for Input Signal Checking in Raspberry Pi Figure 6.1: Icon Figure 6.2: Parent Window Figure 6.3: Overall Structure of the Fault-monitoring Program Figure 6.4: "Connection" Tab Figure 6.5: Monitoring Window. Figure 6.6: Flow Chart for the Fault Detector and Fault Locator Figure 6.7: Fault Location Calculator Figure 6.8: Warning Window Figure 6.9: Screenshot of the Visual Studio 2012 IDE Figure 6.10: Plot of Data Package 285 during Fault Condition Figure 6.11: Phasor Diagram from the Monitor Window during the Fault Condition Figure 6.12: Interface for the Transmission Line Database Figure 6.13: Line to Line and Line to Ground Fault in the Simulation Model viii

9 List of Tables Table 1.1: Specifications Table 4.1: Characteristics of Impedance-based Method and Phasor-based Method Table 5.1: Microcontroller Comparison Table 6.1: Result from the Fault-monitoring Program during Phase A and B Line to Line Fault Table 6.2: Result from the Fault-monitoring Program during Phase B and C Line to Line Fault Table 6.3: Result from the Fault-monitoring Program during Phase A and C Line to Line Fault Table 6.4: Result from the Fault-monitoring Program during Phase A, B and C Line to Line Fault Table 6.5: Result from the Fault-monitoring Program during Phase A Line to Ground Fault Table 6.6: Result from the Fault-monitoring Program during Phase B Line to Ground Fault Table 6.7: Result from the Fault-monitoring Program during Phase C Line to Ground Fault Table 6.8: Result from the Fault-monitoring Program during Phase A and B Line to Ground Fault Table 6.9: Result from the Fault-monitoring Program during Phase A and B Line to Ground Fault Table 6.10: Result from the Fault-monitoring Program during Phase A and C Line to Ground Fault Table 6.11: Result from the Fault-monitoring Program during three Phases to Ground Fault ix

10 Table of Contents Abstract... ii Contributions... iii Acknowledgements... iv Nomenclature... v List of Figures... vii List of Tables... ix Table of Contents... x Chapter 1 Introduction Motivation Project Overview Specifications... 3 Chapter 2 Power System Model and Hardware Configuration Introduction Power System Expansion Load Flow Analysis Power System Simulation Synchrophasor Network Hardware PMU Configuration Current and Voltage Inputs RTDS/PMU Interface Analog Output Scaling Synchrophasor Data Transmission Chapter Summary Chapter 3 Fault Detection Algorithm Introduction Algorithms Current Magnitude Threshold and Phase Selection Method Current Scalar Summation and Vector Summation Comparison Method Test Result and Recommendations Chapter Summary x

11 Chapter 4 Fault Location Algorithm Introduction Accuracy of Fault Location Phasor-based Method vs. Impedance-based Method Principle of Phasor-based Method Experiment Process RSCAD PMU Model and CT Saturation Test Post Fault Condition Measurement Fault Location Calculation Estimated Result and Analysis Various Fault Types Various Fault Impedances Chapter Summary Chapter 5 Backup Protection Implementation Introduction Background Purpose Methods Alternative Logic Circuit Microcontroller Interface Selection of Method Implementation Microcontroller Selection Scheme and Procedure Result Further Improvement Chapter Summary Chapter 6 Windows Application Introduction Structure of the Program Fault Detector Fault Location Calculator xi

12 6.2.3 Fault Alarm Message Graphic User Interface, Plot and Database Estimated Results Line to Line Fault Line to Ground Fault Three Phases to Ground Fault Future Work Chapter Summary Chapter 7 Conclusion References Vita Yaojie,Cai KeKe, Jin Grace,Ngantian Jiajing,Wang Appendix A Appendix B Appendix C Appendix D Appendix E xii

13 Chapter 1 Introduction This chapter contains the motivation for this project, as well as the project overview, followed by the design specifications. 1.1 Motivation Currently, there is a high demand for electrical energy and it is therefore crucial to consider reliability in power systems. Protection systems play an important role in reliability, as they are used to maintain power system stability if any faults occur. These protection systems are used to detect faults when they occur, and then immediately isolate the faults in order to minimize grid damage and prevent power outages. There are various methods of fault detection, but this project focuses on wide area monitoring, which is a very useful tool in protection of large transmission lines. The main cause of large blackouts today is the implementation of poor line monitoring methods that are unable to observe the state of a power system over a wide area. As a result, various instabilities in the network can have a cascade effect over a large distance if they are not detected and no actions are taken to control them. The main objective is to design a differential protection algorithm using synchronized phasor measurements obtained from two ends of a transmission line, and to implement it on real time hardware for testing. The motivation of this project is to detect various types of faults and to determine the fault location on a transmission line, while considering both speed and accuracy optimization. 1.2 Project Overview The main components of this project and the connections between them are shown in Figure 1.1 below. They include: two phasor-measurement units (PMUs) which monitor two ends of a transmission line, as well as a phasor data concentrator (PDC) which will collect synchrophasor data from the PMUs and then send the data to an endpoint computer for fault analysis. These data transmissions will occur through Ethernet. A windows application will be developed to display the synchrophasor measurements and it will also alert the user when a fault occurs. Note that the power system outputs are simulated using the RTDS unit. 1

14 Figure 1.1: Main Setup 2

15 1.1 Specifications Table 1.1: Specifications Specification of Self-developed Window Application Specifications Operating System Method of Communication Data Transmission Rate Protocol Standard Microsoft Windows Client-server 60 Frames per second IEEE Standard for Sychrophasor Data Transfer Stability 1 >99.9% Reliability 2 >99.9% Fault Location Error 3 <5% 1 The probability of Self-developed Window Application not to crash during monitoring process. 2 The probability of Fault Detection function in Self-developed Window Application is not to be triggered by external faults. 3 Error of result of Fault Locator in Self-developed Window Application 3

16 Chapter 2 Power System Model and Hardware Configuration 2.1 Introduction The design of the power system model for this project was done with the intent of testing a very simple system for demonstration purposes, since an intricate system was unnecessary. Also, we can expect similar results for large scale networks because the spacing of the PMUs along the line will approximately be the same. The hardware configurations of the PMUs and RTDS play a large role in this project, since the simulation of a power system must be done with minimal errors in order to generate proper analog outputs that we would expect from a real system. Moreover, the results of the fault analysis algorithms heavily rely on the accuracy of the synchrophasor measurements taken by the hardware PMUs. 2.2 Power System Expansion The preliminary power system model is shown in Figure 2.1. It includes a generator, a transformer, one 210 kv and 207 km transmission line, along with an active load. The generator side will be referred to as Station A, and the active load side will be referred to as Station B. Also, two PMUs are connected at both ends of the transmission line to provide phasor measurements of voltage and current at each station. All of the parameters for the generator, transmission line, transformer were taken from a reliable source to mimic a real case scenario, and are shown in Appendix A [1]. Similarly, the parameters for the instrument transformers can be found in Appendix B. 4

17 Figure 2.1: Preliminary power system model in RSCAD The secondary model shown in Figure 2.2 is an advanced expanded version of the preliminary model. It contains another generator, transformer and transmission line connected to the active load for simplicity. The purpose of using this model was to test the fault detection algorithm on external faults, which are any faults that may occur outside the area contained between the two PMUs. If an external fault happens, the synchrophasor measurements along the internal portion of the power system will differ from measurements taken under normal operation [2]. The fault detection algorithm should ignore these differences because it is unnecessary to trip the breakers along the main line. Figure 2.2: Secondary power system model in RSCAD 5

18 2.2.1 Load Flow Analysis Before simulating the power system, it was necessary to perform a load flow analysis on the proposed design to ensure that there were no errors that may interfere with the experimental results. This was done using PSS/E software, which is commonly used to analyze power system performances. Figure 2.3 shows the drafted secondary model in PSS/E under a load flow analysis. The green arrows show the direction of real power while the orange arrows show the direction of reactive power. Clearly, the model feasible since there are no problems with the load flows power is being supplied by the generators and absorbed by the load. Figure 2.3: Secondary Model in PSS/E 6

19 2.2.2 Power System Simulation The designed power systems were simulated using RSCAD which is compatible with the RTDS. Once the RSCAD model has been compiled, it will allow the RTDS to generate analog outputs in real time that are suitable for the PMUs [3]. These analog outputs include the voltage and current outputs of potential and current transformers along the sending and receiving ends of the transmission line. 2.3 Synchrophasor Network The SEL-421 is a Schweitzer Engineering Laboratories (SEL) relay device that has a wide variety of functions pertaining to protection, automation and controls. Specifically for this project, two of these relays are used as PMUs. They can obtain highly accurate synchrophasor measurements when a GPS clock is connected to these devices. The GPS clock will ensure that the voltage and current measurements taken by both PMUs are synchronized. It is crucial that the measurements are taken at exactly the same time because any time delay between these measurements can impact the results in such a way that a fault may be falsely detected; the algorithm will only be effective when measurements are fully synchronized Hardware PMU Configuration In order to configure the PMUs, there are many parameters that must be set in order to achieve accurate results in synchrophasor measurements. These include global settings, transmission line parameters, and line configuration parameters. The AcSELerator Quickset software developed by Schweitzer Engineering Laboratories Inc. was used to configure these settings. Firstly, the global settings are shown in Appendix C.1 and they are used to set up the main parameters such as location and nominal frequency, while also being used to enable the synchrophasor measurement function. Appendix C.2 displays the transmission line parameters for the transmission line simulated in RSCAD. The positive and zero sequence impedances for the transmission line must be specified to accurately determine the voltage and current values at the sending end and receiving end of the line. 7

20 The relay impedance settings as shown in Appendix C.3 are specified in secondary ohms. These settings are used to find the reach settings of the phase and ground distance elements, and were calculated considering the CT and PT ratios. These calculations are then used to determine the line configuration parameters, which are found in Appendix C Current and Voltage Inputs The PMUs are designed to receive current and voltage signals in analog form, as they are connected to Current Transformers (CTs) and Potential Transformers (PTs) that step down the AC currents and voltages of the transmission line in practice. These current and voltage inputs are shown in Figure 2.4. The nominal current ratings are 1 A or 5 A while the nominal voltage ratings are 69 V. Current Inputs Voltage Inputs Figure 2.4: Current and Voltage Inputs of the SEL-421 The SEL-421 also has the capability of receiving inputs through a low-level analog interface. This interface is typically used for testing in laboratories when the analog signals are simulated, since the ratings are quite low. The nominal current rating is 67 ma while the nominal voltage rating is 446 mv. 2.4 RTDS/PMU Interface Since the PMUs require analog inputs, the digital outputs of the RTDS corresponding to the simulated power system model in RSCAD must be converted to analog form. Therefore, an analog output card was needed to perform the conversion before connecting to the PMUs [4]. The card provided with the RTDS is called a Gigabit Tranceiver Analog Output (GTAO) card as shown in Figure 2.6. The GTAO card is interfaced with the Giga-Processor Card (GPC), which 8

21 solves equations related to the power system in RSCAD in order to simulate the current and voltage outputs. Since the analog outputs of the GTAO card are significantly lower than the outputs of CTs and PTs, these outputs must be amplified before connecting to the PMUs using voltage and current amplifiers [5]. Thus, the gains of the amplifiers can be calculated by using the ratios between the instrument transformer outputs and the GTAO outputs. However, as previously mentioned, the SEL-421 has a low-level interface that accepts analog outputs with lower amplitudes than instrument transformer outputs. Therefore, this project did not require the use of amplifiers. Figure 2.6: GTAO card The basic interface between the RTDS and the PMUs is shown in Figure 2.7. Once the CTs, PTs and CVTs were modeled and compiled in RSCAD, the RTDS will generate the required digital outputs which will then be processed by the GTAO card. Note that the instrument transformer turn ratios were determined by considering the ratio between the current and voltage ratings required by the PMUs and the current and voltage ratings of the transmission line. Instrument Transformers (CTs, PTs, & CVTs) Digital GTAO Card Analog PMUs (SEL-421) Figure 2.7: RTDS/PMU Interface 9

22 2.4.1 Analog Output Scaling The GTAO card must be configured to generate the correct analog signal levels before directly connecting to the low-level interface of the SEL-421 relays. This is done by setting the scaling factor for the card in RSCAD. The equation (2.1) below was extracted from the RTDS manual [1] and it is used to calculate the scaling factors shown in (2.2) and (2.3). (2.1) Current scaling factor calculation: (2.2) Voltage scaling factor calculation: (2.3) Synchrophasor Data Transmission The SEL-421 relays measure the phasor values at both ends of the transmission line synchronously, and will then transmit the synchrophasor data to the PDC, which collects and packages the data. The AcSELerator Quickset software was used to configure the communications for the relays. This allows the user to set up the serial port and Ethernet port parameters. For this project, the synchrophasor data transmission occurs via Transmission Communication Protocol/ Internet Protocol (TCP)/ (IP), although User Datagram Protocol (UDP) is another option. All of the communication configurations are shown in Appendix D. 2.5 Chapter Summary To summarize the hardware portion of the project related to the generation and measurement of synchrophasors, once the power system model was designed and drafted in RSCAD, the RTDS unit collects the model data to produce current and voltage outputs in digital form. The GTAO card is used to convert the digital signals to analog outputs suitable for the low- 10

23 level interface of the PMUs. Afterwards, the PMUs are configured to send the synchrophasor data to the PDC through a TCP network. The PDC will then send the compressed data to an endpoint computer for fault analysis, which is discussed in the next two chapters. 11

24 Chapter 3 Fault Detection Algorithm 3.1 Introduction The fault detection algorithm is based on the Current Differential Protection Method. In the normal condition, a three-phase power system is balanced which means the magnitudes of the currents of all three phases (phases A, B and C) are identical and each phase is separated by a phase angle of 120. During a fault, the transmission system becomes unbalanced, both magnitudes and phase angles of the currents will be changed. By recognizing the changes in the phase currents, the fault itself and the type of it can be detected. 3.2 Algorithms The fault conditions are detected by monitoring the phase impedances, phase-current amplitudes, phase-voltage amplitudes and zero-sequence current amplitudes. The general structure of the fault-detection algorithm is to utilize two different statistic measurements of the voltage or current signals, which are determined with the voltmeter and ampere meter. The criterion value is formulated to assess the fault inception. Based on the Kirchhoff's Current Law, the algebraic sum of all branch currents flowing into any node must be zero. Under normal operation, the measurement of current difference between the two ends of a transmission line is almost zero. However, during the fault conditions, a large difference should exist in the currents [7] Current Magnitude Threshold and Phase Selection Method Fault can be simply detected by setting a threshold value for the transmission line. The threshold value will be the difference of current magnitudes between the two ends of the transmission line under the normal operation condition. The program will constantly compare the present current difference to the threshold value, if a large difference is detected in the comparison; a fault exists in the transmission line. To use this method, the current difference in each phase must be considered. 12

25 This method is straightforward to be developed into a program because of easy coding. The fault detection program will have three "if" statements to compare each phase value of the transmission line. "Present current difference is smaller than the threshold value" will be the condition of the "if" statement, any false condition in the "if" statement will trig the fault alarm. If using the Big O notation (The letter O is used to represent the rate of growth of a function) to represent the length of the process time, the fault detection program can be represented as, which indicates the program has an extremely small increment of growth. Therefore, this method can monitor a large amount of transmission lines at the same time. However, the weakness of this method is the detection of fault types. During the fault conditions, each type of faults will result in a different increment in the current difference. Therefore this method can detect fault types by distinguishing the increment in current difference. In order to prevent errors in fault type detection, multiple threshold values can be set in the program for identifying Double Phase to Ground Fault and Line to Line Fault. Nonetheless, in practice, the setting of multiple threshold values can be extremely complex which would produce a misleading result. For example, a phase A and phase B line to ground fault is simulated in the RTDS. Both phase s current is increasing to a high level, which trigs the fault alarm. However, the warning window indicates it is a phase A line to ground fault as Figure 3.1. Figure 3.1: The false Warning Window for phase A and B line to ground fault 13

26 Figure 3.2: Current vs Package Index Plot for phase A and B line to ground fault As the red circle in Figure 3.2 shows, the increase rate of current magnitude in Phase A and Phase B are not identical, the reason for this phenomenon is the initial rate of change in current contains a large amount of oscillations. The threshold method will tell the fault is on the phase whichever reaches the threshold value in the first place. In this case, the blue line (Phase A) starts to increase before red line(phase B), therefore the incorrect result will be Phase A to Ground Fault instead of Double Phase( Phase A & Phase B) to Ground Fault. However if the user requires the program to determine the correct fault type, an additional function, which can be either Moving Window Function or Phase Selection Function, need to be embedded into this method. Moving Window Function is defined as a program with the ability of collecting multiple times of the current magnitudes and comparing each current measurement to the threshold value. The decision will be made based on all the magnitude comparison results, rather than on one single point of data, in this way, this function will significantly minimize the error. Nevertheless, the Moving Window requires to measure multiple times of data which causes a delay in the program so that the decision may not be made instantaneously. For example, the data transmission rate is one frame per cycle, in order to get accurate result, this function will require five or six cycles to work, but a real system requires much shorter time to response. In addition, the Moving Window Function requires the program to have a nest of "if" and "for" statements, the big O notation for this function will be. The growth rate of the function is increased significantly with the number of the monitoring transmission line, thus this function will require a powerful processor for data processing. 14

27 The Phase Selection Function [3] is a function that identifies the type of a fault by comparing the angles and magnitudes of a sequence of currents. The phase angle between each sequence component and the difference among line currents in phase A, B and C are identified to detect the faulty phase [8]. The currents of the sending end and receiving end in each phase will be added to get a total current value for the transmission line of that phase, as expression (3.8), (3.9), and (3.10). (3.8) (3.9) The phase current difference can be calculated as expression (3.11), (3.12), and (3. 13). (3.10) (3.11) (3.12) (3.13) [ ] [ ] [ ] (3.14) ( ) The sequence component of post fault current can be calculated as expression (3.14). The relationships between the faulty phases are shown in the following expressions (3.15, 3.16 and 3.17) A - Phase Fault: If and (3.15) B - Phase Fault: If and 15

28 (3.16) C - Phase Fault: If ( ) and (3.17) By cross checking the magnitude and phase angle in the sequence component, the fault detection program will be able to identify the type of a fault. In addition, this checking process can be done by using three if statements, the Big O notation to represent the process can be representing as. Therefore the Phase Selection Method is much quicker than the Moving Window Method Current Scalar Summation and Vector Summation Comparison Method Another approach for fault detection is to calculate the scalar and the vector summations of each phase of the transmission line. The fault detection program will compare the result to the Scalar Summation or the Vector Summation to detect the fault [9]. The criterion of Scalar Summation as the bias current can be written as expression (3.18). (3.18) "K" is the restraint coefficient. In order to calculate the restraint coefficient, the sending end current is assuming to be equal to the receiving end current during the normal operation as expression (3.19) (3.19a) (3.19b) From (3.18), the restraint coefficient can be calculated as shown in expression (3.20) (3.20) Plug (3.19) into (3.20), the restraint coefficient will be (3.21) 16

29 The criterion of Vector Summation as the bias current can be written as expression (3.22) (3.22) From (3.22), the restraint coefficient can be calculated as shown in expression (3.23) (3.23) Plug (3.19) into (3.23), the restraint coefficient will be (3.24) The current scalar comparison is applied in each phase of the transmission line. This method is better than the current threshold method because this method is taking account of both the magnitude and phase angle. The Scalar Summation and Vector Summation Method is not rely on only the comparison of the magnitude difference between the sending end and receiving end currents to the threshold value. Besides that, for the Current Magnitude Threshold and Phase Selection Method, the fault detection program will have three "if" statements to compare the value in the each phase of the lines. The Big O notation to represent the process can be represented as Test Result and Recommendations During the lab test both methods showed high performance with 100 percent accuracy during different fault conditions. The result of the testing is discussed in the Chapter 6 Windows Application. Moreover, time-consuming is another important standard to evaluate the fault detection method. Figure 3.3 and Figure 3.4 will show the length of time that each method need to analyze each data package from data package seventeen to data package twenty-three. 17

30 Figure 3.3: Time for Current Magnitude Threshold and Phase Selection Method to Detect Fault Figure 3.4: Time for Current Scalar Summation Method to Detect Fault According to C IEEE Standard for Synchrophasor Data Transfer for Power Systems, the highest rate for the data transfer is 200 frames per second which is seconds between each data package. The average time for Current Magnitude Threshold and Phase Selection Method to detect fault is secends. The average time for Current Scalar Summation Method to detect fault is seconds. Both methods are shorter than seconds. Current Scalar Summation Method takes a much shorter analysis time than the Current Magnitude Threshold and Phase Selection Method, this advantage will allow the program to provide more real time analysis (such as transmission line parameters calculation) for the data package. 18

31 3.3 Chapter Summary In this chapter, two fault detection methods were presented, which are the Current Magnitude Threshold and Phase Selection Method and the Current Scalar Summation and Vector Summation Comparison Method. A number of tests were performed to validate the above method. The test results and the recommendations were discussed in detail. The Current Scalar Summation and Vector Summation Comparison Method has been chosen to implement the fault detection system in this project. 19

32 Chapter 4 Fault Location Algorithm 4.1 Introduction The calculation of transmission line fault location has been the primary subject to power system protection study. The main reason is transmission-line-faults occur more often than faults in sub-transmission and distribution system; approximately two third of faults which occur in power system belong to transmission-line-faults [10]. Additionally, the major-transmission line is usually above 100 km and the crossing area of transmission line is normally various terrains which maybe river, forest and mountain, therefore the patrol time required to pinpoint fault location of transmission lines is much longer than the time spent in sub-transmission and distribution system. For these reasons, developing the Fault Locators in transmission line; which can help accurately identify locations for early repairs, have been received extensive attention [11]. The accuracy of fault locator is great importance to economic operation of transmission lines. Transmission line operation cost can be largely reduced because the accurate fault location can avoid lengthy and expensive patrol time; however, a small measurement error may cause detailed local examination over several kilometers of transmission line. Once the accurate fault location has been pinpointed, the transmission line can be restored without delay, the timely repairs can prevent major damages to power system, ultimately reducing revenue loss caused by outages. In order to pinpoint the fault location accurately, the factor which will influence fault location accuracy need to be study, those factors is detailed discussed in section 4.2. The fault location algorithm based on synchronized measurement can be generally classified two categories: Impedance-based Method and Phasor-based Method [2]. Theoretically, Phasor-based method is more accurate than Impedance-based Method, consequently, this chapter experiment more focus on verifying the accuracy of Phased-based method. The detailed comparison of these two algorithms will be discussed in section 4.3. The principle of selected algorithm (Phasor-based Method) for a single phase transmission line which then is utilized as prototype formula, is derived in section 4.4. Finally, the simulation and associated result are presented in section

33 4.2 Accuracy of Fault Location The fault-location error is defined as following expression: Percentage error in fault-location estimate based on the total line length: (error) = (calculated distance exact distance to the fault) divided by (total line length) This definition can be written down as the following formula: Error (%) = (4.1) where - estimated distance to the fault; dexact exact distance to the fault; Total line length The units of above three parameters can be in km or in relative per unit (p.u.). The purpose of study is to obtain the average error of Phasor-based Method for a given population of the evaluation tests; however, the errors having identical magnitude but different signs do not compensate each other; therefore it is characteristic that the absolute value is usually taken for the nominator [2]. In general, without specifying the fault-location method, when performing the faultlocation accuracy evaluation, different factors affecting the accuracy are taken into account. The main source of error can be listed as follows: 1. Inaccurate line parameters. Line parameter in model do not exactly equal to the actual parameters. Even if the geometry and type of line conductors are accurately taken for calculating the line impedances, the total line length is difficult to accurately estimate. 2. Unknown fault impedance. Fault impedance may be described as an unpredictable quantity, because it maybe an electric arc, tower grounding, or the presence of objects in the fault path. 3. Various remote source impedance. Switching operation of several generator on the remote side of transmission network, will change the source operation form or source impedance from those assumed setting of power system model, which in turn can be a major error in fault location. 21

34 4. Inaccurate fault-type identification. Weak source conditions challenge the one-ended fault locators, because the system is more likely nonhomogeneous which make fault type selection more difficult [12]. 5. Neglecting the presence of compensating devices. Under high resistance fault conditions, capacitance current for long lines can be comparable with the current in the fault path [2]. 4.3 Phasor-based Method vs. Impedance-based Method In order to reduce or eliminate the factors which will influences the accuracy of fault location, Phasor-based Method and Impedance-based Method was analyzed and compared in this section. Phasor-based technique uses synchronized-phasor measurement which identify post-fault voltage and current at both line ends. Fault location is independent of fault resistance and the method does not require any knowledge of source impedance. It maintains high accuracy for untransposed lines and no fault type identification is required [10]. For these reasons, many of the fundamental limitations on the accuracy which has been listed in section 4.2 achievable are reduced. Impedance-based Method is famous for its One-ended measurement techniques which includes Simple Reactance Method, Takagi Method and Modified Takagi Method. A major advantage for these techniques is that communication channel and remote data are not required [12]; therefore, the error caused by various remote source impedances can be eliminated. Only simple implementation into digital protective relays or digital fault recorders will be feasible. However; several factors will dramatically bring down the accuracy of Impedance-based Method. Firstly, Impedance-based Method ignores the effect of fault impedance; in power system study, the fault impedance is major factor to affect the calculated fault location result. Secondly, this method is only suitable to an ideal homogeneous system, more unlikely inhomogeneous system is, and the larger error in the fault location will exist. Thirdly, Impedance-based Method is largely depends on the fault type identification, failed to identify the correct fault type will cause the error in estimated fault location. Table 1 shows the characteristics of two method discussed above. Comparing these two methods, advantages and disadvantages of them can be found. 22

35 Table 4.1: Characteristics of Impedance-based Method and Phasor-based Method Impedance based Method Phasor based Method Pros Cons Pros Cons 1. Sometimes only require one ended T line measurements 2. Communication means are not needed 1. Ignoring the effect of fault impedance 2. Slow- speed computing time 1. Independent of fault resistance, source impedance. 2. Suitable for untransposed transmission line 3. No fault type identification required Require synchronized-phasor measurement and communication devices To improve the fault-location estimation, it is important to eliminate, or at least to reduce possible errors for the considered method. From analysis above, the Phasor-based Method is more accurate by offering improved fault-location determination without any assumptions of fault impedances and information regarding the external networks such as impedances of the equivalent sources. In this way, Phasor-based Method is decided to apply to the simulation study. 23

36 4.4 Principle of Phasor-based Method Consider the transmission long line model with length l which is shown Figure 4.1 Figure 4.1: Transmission Long Line Model X = distance to fault from receiving end, = Voltages at fault, sending and receiving ends, = Currents at sending and receiving ends L = line length As Δx 0, (4.2) (4.3) Differentiating (4.2) and substituting from (4.3) (4.4) Where γ= which is known as propagation constant The solution of (4.4) is: (4.5) 24

37 To find the constant, we set x=0, and, (4.6) (4.7) Where = which is known as characteristic impedance Substitute (4.6) and (4.7) into (4.5), the hyperbolic function with sinh and cosh can be obtained: (4.8) Using the same principle, deriving the equation from the sending end, (4.9) According to hyperbolic function properties, (4.10) (4.11) Substitute (4.10) and (4.11) into (4.9) and organized the equation, (4.12) Rearrange the (4.12) into the fault location expression form, (4.13) where X is the distance to fault from receiving end; stands for transmission line length; as described above; is characteristic impedance; γ is known as propagation constant. are post fault voltages at sending and receiving ends respectively; fault currents at sending and receiving ends respectively. are post 25

38 If all parameters above could be identified without error, (4.13) would lead to an exact evaluation of the distance to fault X. Theatrically, X should be calculated as a real value; However, in the simulation case, the calculated value of X has a very small imaginary part which can be ignored; only the real part of calculated value is thus taken to represent the fault distance. In equation (4.13), impedance and admittance is known based on the parameters of transmission line, the post fault voltages and currents are measured by PMU. Additionally, since the formula (4.13) does not contain the source impedance and fault impedance parameters, therefore source impedance and fault impedance are independent of this fault algorithm. Lastly, Phasor-based is derived from the even distributed line, and inherently includes the effect of compensating devices such as the shunt capacitance. 4.5 Experiment Process Phasor-based Method has been determined by implementing RTDS simulation to verify the accuracy of algorithm. PMU Model was created in the RSCAD network (network configuration detail see 2.1.3), and all the fault location results are produced based on the measurement of PMU in RSCAD. One of the most important feature of PMU in RSCAD is simulating the real time environment for SEL-421(Digital Relay). With assist help from CT in the model, PMU can monitor the current so that adjusting CT into non-saturation state during the fault condition. Instruction of RSCAD and CT saturation test is presented in The measurement technique obeys the principle of synchronized phasor measurement, post fault voltage and current are selected to be at exactly same time after fault occurs. The detail information of post fault measurement will be discussed in section The purpose of software calculation which includes MATLAB and C# (see section 4.5.3) is ensure the accuracy of Phasor-based Algorithm so that implement an accurate algorithm into hardware PMU (Digital Relay); therefore, performance evaluation in this chapter reveals the inaccuracies in the measurement form PMU Model in RSCAD itself, and does not include any errors caused by the PMU hardware. Section 4.6 presents the calculated fault locations under different fault types and fault impedances along the transmission line. Three Phase fault (3 Phase Fault), Single Line to Ground(LG)and Double Line to Ground (LLG Fault) was investigated at different locations of transmission line. Fault impedance were assumed to be 10 p.u. and 20 p.u., so as to identify the effect of fault impedance on the accuracy of the Impedance-based Method. Based on the two conditions (different fault types and different fault impedances), the fault was 26

39 set on every five percent of transmission line to compare the error percentage (equation 4.1) between the estimated result to actual fault location RSCAD PMU Model and CT Saturation Test Phasor Measurement Unit (PMU) models including P Class, M Class and 16 point DFT with PLL tracking. In this study, two AnnexC P class PMU was configured at both end of Transmission line. For simulation purpose, the reporting rate of PMU was select 60 frames per second which is exactly same to the specification rate of Windows Application (Table 1.1). The voltage and current turns radio was adjusted to match the setting of CT and PT; respectively 240 and 230. In order to obtain the correct current measurements, performing the CT saturation test before conducting fault analysis is extreme necessary. Typically, current transformers are designed to operate well within the linear region of the flux current plane. However, under heavy fault conditions the current transformer may saturate, which lead to underestimate the post fault current. Before sampling the post fault current, the Turns Radios need to be adjusted until the CT is not saturated. According to Turns Radios of CT, the estimated primary current can be calculated by following formula: Estimated primary current = Actual Secondary Current Turns Radio (4.14) Secondary side clarified as low side of transformer and Primary side is noted as high side of transformer. Figure 4.2 shows an example of when CT is not saturated during the fault condtion which shows the estimated current (IAs) in red line and actual current (IBRKA2) in green line which are overlapping to each other. Since the estimated primary current is equal to actual measured current even the fault condition, the CT is not saturated. 27

40 Estimated Current (ka) Actual Current (ka) Time(secs) Figure 4.2: Unsaturation Case of Estimated Current (IAs) vs Actual Current (IBRKA2) Post Fault Condition Measurement Accurately identified the Post Fault Conditions which includes post fault voltages and currents is great importance to pinpoint fault location. LG (Phase A to Ground) fault was simulated at 90 percent length from sending end of transmission line. Figure 4.3 demonstrates the voltage curve of transmission line sending end after fault occurred. As it can been seen on the Figure 4.3, voltage reaches the lowest point 41.5 kv at 0.35 seconds, that point will be the post fault voltage which was used section for calculation the fault location (Vs). Utilized the similar principle, the receiving end of post fault voltage can be found (Vr). 28

41 Figure 4.3: Post Fault Voltage Curve In Figure 4.4, the post current is identified as current on the plot with the exactly same time of the post fault voltage, since the synchro-phasor measurement always measures the voltages and currents at same time. In this example, at 0.35 seconds, the current value is 3126 amps, this value will be implemented in the fault location as post fault current. Both values was determined by PMU in RSCAD by using the cursor tool to be precisely measured. 29

42 Figure 4.4: Post Fault Current Curve However, in order to measure the all three phase voltages and currents with phase and magnitude, it is earlier to measure the real and imaginary parts instead of magnitude and phase parts, because RSCAD could not generate the transient waveform of angle changes of the voltages and currents. Figure 4.5 and Figure 4.6 shows the real parts and imaginary parts of the post fault voltage which is convertible to the magnitude and phase parts. 30

43 Figure 4.5: Real Parts of Voltage Phase A on the Sending End of Transmission Line Figure 4.6: Imaginary Parts of Voltage Phase A on the Sending End of Transmission Line 31

44 4.5.3 Fault Location Calculation Matlab and C# program was implemented to calculate the fault location under the given parameters from RSCAD network such as line impedance and admittance. From known parameters, the surge impedance and propagation constant can be identified. The post fault voltages and current are measured by PMU Model in RSCAD (4.5.2). Eventually, the equation (4.13) was utilized to produce the fault location result. The flowchart of Matlab code is shown in Figure 4.7. Identifying Parameters from RSCAD Model (l, z,y) Identifying Parameters from RSCAD Model (l, z,y) Identifying Parameters from RSCAD Model (l, z,y) Fault Location 1 γ tan 1 Z CI S sinh lγ V R V S cosh lγ Z C I S cosh lγ V S sinh lγ Z C I R Figure 4.7: Flowchart of Matlab program for calculating the Fault location 4.6 Estimated Result and Analysis This section presents the influences of fault impedance and fault type on the measured fault locations which predicts the possible outcomes when implemented of digital relay. 3 Phase Fault, LG Fault and LLG Fault was investigated at different locations of transmission line. Two different fault impedance were assumed under LG Fault condition. The purpose of it is verifying the Phasor-based Method is independent of fault impedance. The program which calculates fault 32

45 location (MATLAB) is set using the exact parameters from the RSCAD network model, the only errors that occur are those due to measurement error of post fault voltages and currents from PMU Model. In other words, the algorithm error should be zero Various Fault Types Figure 4.8 shows the absolute value of error percentage of calculated fault location of 3 Phase Balance Fault condition. The investigated fault location was ranged from 10 % to 90 % of transmission line with every 5% step change. From Figure 4.8, the curve of error forms into symmetrical parabola with opening upwards. The largest error occur at sending and receiving end (10% and 90% actual fault location), which numerically 0.88 %. The most accurate location was identified at middle of transmission line (50% percent of actual position), the accuracy reaches to 0.05 %. 0.90% 0.80% 0.70% Measured Error% 0.60% 0.50% 0.40% 0.30% 0.20% 0.10% 0.00% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Actual Fault Postition (%) Figure 4.8: 3 Phase Fault -Measured Error percentage vs Actual Fault location Figure 4.9 illustrates the accuracy of Phasor-based Algorithm under LG faults condition. Error percentage of calculated fault location dramatically increases to 2.25% at both ends of transmission line, which is three times larger than the result under 3 Phase Fault condition; nevertheless, the error maintains below optimistic level from 25% to 30% location of transmission line. 33

46 2.50% 2.00% Measured Error% 1.50% 1.00% 0.50% 0.00% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Actual Fault Postition (%) Figure 4.9: LG Fault -Measured Error percentage vs Actual Fault location The worst scenarios of Phasor-based Method is the occurrence of LLG Fault condition, the result can be seen from Figure The measured error is as high as 2.55%. Additionally, the overall error is brought up except for the middle point of transmission line. 3.00% 2.50% Measured Error% 2.00% 1.50% 1.00% 0.50% 0.00% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Actual Fault Postition (%) Figure 4.10: LLG Fault -Measured Error percentage vs Actual Fault location The result of three different types of fault presented in shows that Phasor-based Method gives an accurate evaluation of fault position that is independent of the actual fault point, 34

47 even the measured error percentage has the difference between the mid-point and terminal ends of transmission line, but the largest difference is only 1.5% which can be ignored. In other words, the algorithm predicts the slightly more accurate result when the fault is set in the middle range of transmission line than when the fault is closed to both terminals of transmission line, however, overall accuracy can be generally treated as uniform. The accuracy for pure phase faults clear of ground (3 Phase Fault) has been found to be better than that for earth faults (LLG and LG Faults). Among three types of fault. For reasons of conservative study, only results for LG Fault condition are given in Various Fault Impedance; since LG Fault provides the medium accuracy result Various Fault Impedance LG fault is selected as a simulation case whose fault impedance are 10 p.u. and 20 p.u. and the result shows in Figure From 20 percent of 80 percent of transmission line, the difference of measured error percentage for both case is within 0.5%, which means Phasor-based Method is generally independent of fault impedance. Exceptional case is the occurrence of fault closed to terminals of transmission line. The extreme case is found at fault location at 15% and 85% of transmission line. Figure 4.11: Various Fault Impedance -Measured Error percentage vs Actual Fault location 35

48 4.7 Chapter Summary A relative accurate PMU-based fault location algorithm (Phasor-based Method) for transmission lines is presented in this chapter. Phasor-based Method is evaluated under three different fault types, the difference of error percentage among them is average 0.95%. With adding two fault impedances, two average errors show significantly closed to each other which are 0.96% and 0.91%. Furthermore, the overall average error at different points of the transmission line is 1.1% error. Therefore the simulation studies demonstrate that the Phasorbased Method is not significantly affected by various system situations such as fault types, fault impedances and fault locations. Even some variation in accuracy exists in this study, but under all conditions a high degree of accuracy has been achieved. With the advancement of digital relay technology, Phasor-based Method is very feasible and effective for calculating the accurate fault location. 36

49 Chapter 5 Backup Protection Implementation 5.1 Introduction The main objective of a backup protection is to open all the connections of generation and load with the transmission line to an uncleared fault. In order to perform this objective and to obtain a high reliability of a transmission system, the backup protection system in this project will be able to meet the requirements for implementing breaker functions [13]: 1. It must recognize the existence of all faults that occur on the transmission line. 2. It must detect the failure of the primary protection system. 3. It must distinct internal and external faults associated with a specific transmission line, and trips the minimum number of breakers on that transmission line only when internal faults occur. 4. It must operate fast enough to preserve the system stability and to prevent device damages. The main program of this project will accomplish requirements one and three, and both of them were introduced in Chapter 3. The backup protection in this project is used as a secondary source to open the breakers in transmission line, and is operating simultaneously with the primary protection system. Consequently, the second requirement of implementing a backup protection above is not necessary. More detailed information will be discussed in section Background One of the major features of the SEL-421 relay is protection, which is the primary protection system in this project. As shown in Figure 5.1, the SEL-421 contains all the necessary protective elements and control logic to protect transmission lines. 37

50 Figure 5.1: SEL-421 Functional Overview [14] The relay simultaneously measures five zones of phase and ground mho distance plus five zones of phase and ground quadrilateral distance. These distance elements, together with optional high-speed directional and faulted phase selection (HSDPS) and high-speed distance elements, are applied in communications-assisted and step-distance protection schemes [14]. By programming the specific elements, combination of elements and inputs using the SEL control equation, the communications scheme tripping can be performed. The second scheme of achieving the protection is to accomplish a logic circuit in the simulation that processes the specific elements in order to control the breakers. This second method is related to the function Expanded SELogic Control Equations and will be introduced later in section Additionally, the breaker failure protection provides a more flexible operation. This function allows monitoring current individually in two breakers and gives a higher sensitivity for detection of circuit breaker opening. This feature is essential if breaker failure is initiated on all circuit breaker trips. 38

51 5.1.2 Purpose The purpose of implementing a backup protection in this project is to assure that the transmission system can be protected as soon as possible when any type of fault occurs. Normally, backup protections are only operating when the primary protections fail to work. In that case, one of the requirements for a backup protection will be checking the status of the transmission system and able to detect the failure of the primary protection system. On the other hand, as mentioned earlier in this chapter, the backup protection in the project will be operating simultaneously with the primary protection system to ensure that the breakers can be open successfully in the shortest period of time after a fault occurs. By applying this strategy, the detection time can be saved when the primary protection system is out of order. In other words, the protection system of this project will be more reliable and will be able to prevent transmission lines from more possible damages that would happen in the duration of detection time. 5.2 Methods In this project, two different methods have been considered to implement the backup protection function. These methods are alternative logic circuit design and microcontroller interface application, detailed information of the schemes will be introduced in section and section respectively. The microcontroller interface method is selected for this project Alternative Logic Circuit In section 5.1.1, a function of the SEL-421 relay called Expanded SELogic Control Equations has been introduced. This function is one of the built-in protection systems in the relay. The basic principle is to process several specific data measured from the transmission system by an equation that is expressed by a digital logic circuit. Figure 5.2 shows a basic logic circuit that controls the breakers in the device. 39

52 Figure 5.2: Original Breaker Logic Circuit Based on this original logic circuit in the program, the internal breakers can be controlled to open or close by different signals. Initially when there is no fault in the transmission system, the status of the breakers is closed. When a fault occurs on the transmission line, the FLTSEQ signal will change, and after processing this signal by the whole logic circuit, an open signal will be given to the breakers. When the fault is cleared, the FLTSEQ signal will change back to the initial value, and the breakers will be manipulated to close again. The first method to implement the backup protection function is to design an alternative digital logic circuit besides the original circuit in the program. The alternative circuit will apply another set of logic that process a different signal measured from the transmission lines. This secondary circuit will be tested individually and operated simultaneously with the original logic circuit. When a fault happens, both circuits will be running at the same time to ensure that the breakers will be open successfully. In this case, even if the original circuit fails to give a correct instruction to the breakers, the alternative circuit will offer a supplementary command to the protection system. 40

53 5.2.2 Microcontroller Interface A second method of implementing the backup protection function of this project is to apply a microcontroller interface. For accomplishing this method, a microcontroller will be using to communicate with the main program and the breakers. In this method, the microcontroller will consistently examine the signal port and receive data from the main program. The main program will send a signal to the microcontroller when a fault is detected on the transmission lines. After reading the fault detection signal from the main program, the microcontroller will control the breakers to open immediately. In order to protect the transmission system as soon as possible, a feasible shortest program has to be applied to the microcontroller. Figure 5.3 below shows the logic of the completed program for this project. In this project, an LED light will be controlled to flash by an output pin on Raspberry Pi instead of using real breakers for demonstration. Figure 5.3: A complete program for the microcontroller interface method 41

54 In section 5.3, more detailed description of the microcontroller interface method will be provided including selection of a microcontroller, method and procedure of the implementation, logic of the program and demonstration of the backup protection function. An additional improvement to this method is to connect the microcontroller to the RTDS device, and to manipulate the breakers in the relay directly by the microcontroller. By implementing this improvement, a fully complete simulation of the project can be achieved. This enforcement will be discussed in section Selection of Method The method microcontroller interface has been chosen to implement the backup protection system. Since the communication between the main program and the microcontroller is transmitted by internet, therefore by using a microcontroller in this project, remote control can be achieved. The second reason is the transmission system can be more reliable by accomplishing microcontroller interface. A microcontroller interface application is a system totally different with the primary protection. If an alternative digital logic circuit is applied as the backup protection, there will be a risk that both of the protection systems are failed to function. This problem is because both of the protection systems are having exactly the same operating principle. On the other side, when accomplishing the backup protection system by the microcontroller interface method, a completely different system scheme will be processed. In this circumstance, the whole protection system of the project can obtain a true double assurance. Furthermore, a new feature of the RTDS device can be achieved with a completion of the improvement of the microcontroller interface method. 5.3 Implementation In this section, all the information about implementing the backup protection system is discussed in detail. The microcontroller that is used in this project as well as the reasons of choosing it will be introduced. Detailed description of each step of the procedure to implement this backup protection function will also be supplied. 42

55 5.3.1 Microcontroller Selection For choosing a suitable device, several popular microcontrollers in the market have been considered and discussed from different aspects. Table 5.1 below shows a brief comparison of the famous microcontrollers. Table 5.1: Microcontroller Comparison Microcontroller Raspberry Pi Arduino Uno Beagle Bone Speed 700 MHz 16 MHz 700 MHz RAM 256 MB 2 kb 256 MB ROM SD card 32 Kb 4 GB Flash + microsd card USB 2x2.0 N/A 1x2.0 Video HDMI, NTSC, JTAG N/A N/A Internet Interface Yes No Yes Price $35 $39 $89 Raspberry Pi has been selected to implement the microcontroller interface part for this project. By comparing the specification data of each microcontroller provided in the table, apparently Raspberry Pi has huge advantages over the others. Compare with Arduino Uno and Beagle Bone, Raspberry Pi has the highest processing speed and the largest RAM space. Other than those advantages, Raspberry Pi is the only microcontroller has an HDMI connection port, which will provide a much more convenient way to monitor the operation system and interface. The two USB ports on Raspberry Pi even allow connecting a mouse and a keyboard to control the system and create and edit any programs that are required. On the aspect of ROM space, even though Beagle Bone has the largest ROM, Raspberry Pi is still a strong competitor with flexible ROM selections. The most important feature of a microcontroller for this project is the internet interface function. In order to accomplish the completed operation, the microcontroller used in the project 43

56 must have the ability to connect to the internet and to receive signals from a static IP address. In this condition, the main reason that the microcontroller Arduino Uno is not a suitable selection for this project is because Arduino Uno does not have an internet interface function. With regard to the prices of Beagle Bone and Raspberry Pi, Beagle Bone is almost three times as expensive as Raspberry Pi. Nevertheless, both Beagle Bone and Raspberry Pi possess similar and competitive of functions. From all the comparisons above, the microcontroller Raspberry Pi is the most reasonable and suitable device for this project Scheme and Procedure The completed implementation of backup protection function can be divided into several sequential parts: Raspberry Pi operating system installation, demonstration testing, Raspberry Pi internet connection and signal receiving from the main program and the final test that combines all the steps above as a whole. Each step of the procedure will be described in detail in this section. Six different operating systems are supplied on the official website of Raspberry Pi, they are Raspbian, Pidora, RaspBMC, OpenELEC, RISC OS and Arch. The system Raspbian has been chosen, this system is the one recommended by the Foundation and the most widely used. As a result, there are plenty of great sources and supports around the internet for new learners. Another reason of choosing Raspbian is this system is based on Debian, which is an extremely well respected and capable Linux distribution, and hence Raspbian would be the easiest system to start with. After installing the operating system in Raspberry Pi, the second step is to test the demonstration part. An LED light will be using for demonstration instead of real breakers. In this step, a short program created by Python is accomplished in Raspberry Pi, where an output pin that connected to the LED need to be assigned first, followed by a loop instructs the LED to blink. The logic of the program for only the demonstration part is shown in Figure

57 Output pin = 1 Sleep (2) Output pin = 0 Sleep (2) Figure 5.4: Program for the LED demonstration For implementing the third step, a static IP address needs to be assigned in order to connect Raspberry Pi to the internet. Both main program and Raspberry Pi will need to finish setting up the IP address in the programs to obtain the communication between them. The main program will send a signal to Raspberry Pi through the internet by the static IP address while Raspberry Pi is running a program of continually checking the internet connection port for receiving signals. Figure 5.5 shows the working principle of the signal receiving part. This program is created by the language Python, the reason of using this language is because Python code is easier to write, to read and to understand, another main reason is Python is a build-in language in the system Raspbian. 45

58 Signal from Main = true? No Yes LED demonstration Figure 5.5: Program for input signal checking in Raspberry Pi The final step of implementing the completed backup protection system is to combine all the previous steps with separated logic programs together and make a final test. The demonstration program will be embedded into the signal receiving program, once a signal from the main program is detected by Raspberry Pi, the demonstration program will automatically start operating to light up the LED, which means to open the breakers in the transmission system. The logic of the completed system is shown previously in figure 5.3 section Result By following the scheme and the procedure described in section 5.3.2, the backup protection function has been implemented successfully as required for the project. Although some unexpected extra time was taken for applying and building the IP address for Raspberry Pi in the laboratory, the whole part of implementing the backup protection system was able to be completed on time. 5.4 Further Improvement A further improvement of implementing the backup protection function is to connect the microcontroller Raspberry Pi to RTDS in order to control the breakers in the relay. To accomplish this additional function, a GTDI (Giga-Transceiver Digital Input Card) card will be needed for 46

59 the RTDS equipment for receiving external control signals. After connecting Raspberry Pi with RTDS by the GTDI card, Raspberry Pi will be able to communicate with RTDS directly. When the signal of opening breakers is received by Raspberry Pi from the main program, another new signal will be sent out from Raspberry Pi to RTDS through the GTDI card. The new signal will be connected to the breaker function in RTDS, and when the equipment receives the signal, the program will process the signal immediately and open the breakers in the relay. Hence an external control of relay breakers will be built in this improvement. For achieving a more complete simulation and RTDS device control, this improvement will be added in the future work to this project. 5.5 Chapter Summary The principle, method and procedure of the implementation of backup protection function were described in detail in this chapter. The original objective and expectation of this function have been achieved successfully for the project. The additional further improvement introduced above in section 5.4 will also be accomplished later to obtain a more complete backup protection implementation with RTDS. 47

60 Chapter 6 Windows Application 6.1 Introduction The fault detection and location calculation algorithm described in the previous chapters were implemented on a PC as a stand-alone windows application written in C# language. This Fault-monitoring Program is using.net framework and Windows Form (Winform) platform. This Fault-monitoring Program has a size of three hundred mega-bytes with the main program, ILNumerics numerical library and a Microsoft Access database. The development of Faultminitoring program has two stages: First stage program and second stage program. The first stage Fault-monitoring Program was developed in C environment by using Cygwin and Textpad and program which is a console application. In order to execute the code, the user has to memorize the command. In other words, the original program is not user-friendly. The second stage Fault-monitoring Program was implemented by using Visual studio 2012 development tools. Compared to the first stage fault-monitoring program, the second stage program was developed to be more feasible and advantageous. Firstly, the second stage program has an excellent user interface, a clear distribution between the working area and library, and the development tool is free. Secondly, the second stage program has a classic parent and child window structure. The parent and child window structure is accessible for other developer to insert additional functions into the library and even possible to reconstruct the working area to adopt other purposes in the future. Thirdly, second stage Fault-monitoring Program can monitor the status of a transmission system in real time. The program provides user high quality of plot and animation, which allow the user to obtain a better insight during the fault condition. Fourthly, the program can send a fault alarm message to a PC or a microcontroller in the power factory to control the breaker in the each end of the transmission line through the internet web. This communication channel will allowed the power system network to have a quick protection response after the fault occur. In section 6.2, the structure of the Fault-monitoring Program is explained in detail. The test result and the recommendations for future research are covered in section

61 6.2 Structure of the Program As mentioned in the introduction, this Fault-monitoring Program is developed by using the parent and child window structure. A child window is confined to the client area of its parent window, an application typically uses child window to divide the client area of a parent window into functional areas. In the fault-monitoring program, the parent window is the main frame of the program. The parent window is a starting window of the program; as well, this window is an archive for the user to get access to the other functions inside the program. The user can access the parent window by double click the icon inside the program file. Figure 6.1 is a picture for the icon. Figure 6.2 is a picture for the fault-monitoring parent window. Figure 6.1: Icon Figure 6.2: Parent Window 49

62 Figure 6.3: Overall Structure of the Fault-monitoring Program Figure 6.3 illustrates the Fault-monitoring Program with two parts divided inside the parent window. The first part is the working area, which is the monitoring window. The second part is the function library. The monitoring window can be accessed by click the "Connection" tab in the parent window and click "Connect the Phasor Data Concentrators". Figure 6.4 is a picture for the "Connection" tab. Figure 6.4: "Connection" Tab 50

63 Figure 6.5: Monitoring Window. Function library can be accessed by click the "Calculator and Convertor" tab. The data package is received by the program through the internet which is from a TCP connection port. As Figure 6.3 shows, there are five input and output ports in the fault monitoring program. Two of the TCP connection ports for the program are for receiving the data package from the PDC and sending out the fault alarm message to the microcontroller in power factory. One of the input ports for the program is for drawing the transmission line parameter from the transmission line database. Another output port for the program is for storing the information database of the status of the transmission system during a fault. Post fault analysis can be provided by analyzing the voltage and current information in the database in the future. One DLL file will hook the ILNurmerius numerical library to the program, and the ILNurmerius library will help to accelerate the calculation process in the monitoring window. The monitoring window or the working area is the main part of the program. The fault detector and fault location calculator tool can be drawn from the function library in this window. The following sections will explain the major functions in the function library. 51

64 6.2.1 Fault Detector The fault detector does not have user interface, because it needs stream of data to provide fault detection. When the user clicks the "Connection" button on the monitoring window (Figure 6.4), the fault detector is called first. Figure 6.6: Flow Chart for the Fault Detector and Fault Locator The first step is to create a connection between the fault detector and PDC by entering the IP address and the port number of PDC which is provided by a user. The monitoring program will wait 12 seconds to the entire PDC server to response. The fault detector function will return if the connection is not established. When the data package arrived, the fault detector function provides decoding according to C IEEE Standard for Synchrophasor Data Transfer for Power Systems. The data package is separated into twelve polo forms of voltage and current values, and those values are 52

65 processed by using the fault detection algorithm purposed in Chapter 3 in order to detect a fault. If a fault occurs on the transmission line is detected by the fault detector, the fault location calculator will be called immediately to estimate the fault location on the line Fault Location Calculator Figure 6.7: Fault Location Calculator Users can enter a set of post fault data to calculate the fault location by using the fault detection algorithm purposed in the chapter 4. However, during the monitoring in this program, after the fault detector detects a fault in the power transmission line, the detector will call the fault location calculator automatically and inject the post fault data into the fault location calculator. The location result from the calculator is presented in a pop warning window as shown in the Figure

66 Figure 6.8: Warning Window Fault Alarm Message After the warning window executed, the program will return to the monitoring window. The monitoring window acknowledges the fault in the fault detector, and the window will open the TCP connection to the microcontroller with a static IP address and a port number. The fault alarm message will be sent with a communication established between the monitoring window and the microcontroller. The pre-set fault alarm message is "there is a fault" in this program and a same message is stored in the microcontroller as well. The fault alarm message is changeable by user s preference in the future. The microcontroller will compare the received message and the pre-set message to avoid harmful programs connecting the microcontroller and falsely tripping the breaker. If the fault alarm message is matched, the microcontroller will be aware of the fault occurred and then send a tripping signal to the breakers Graphic User Interface, Plot and Database The Graphic user interface is created by using Visual Studio Figure 6.9 is a screenshot of the Visual Studio 2012 IDE. 54

67 Figure 6.9: Screenshot of the Visual Studio 2012 IDE In order to create a plot of the current and voltage magnitudes verse the index number of the data package, the program pushes all the data from the fault detector into a data collection. This data collection is sorted by the index number of the data package, and the collection is updated when a new data package is received. The high frequency of the update creates a real time animation for status of a transmission line to give the user a better visual experience. Figure 6.10: Plot of Data Package 285 During Fault Condition 55

68 The phase angle plot has the same principle of the current and voltage magnitude plot. The monitoring window will convert the data inside the data collection into polo form. The monitoring window creates a small rectangular Cartesian coordinate system in the right hand side of the window. The monitoring window draws a line for each data such as phase A voltage. The starting point of the line is the origin of the coordinate system, the angle of the line is the phase angle of the data, and the length of the line is the magnitude of the data. Figure 6.11: Phasor Diagram from the Monitor Window during the Fault Condition Microsoft Access Database is used to store the transmission line data and the post fault data when a fault occurs. (As shown in Figure 6.12) Figure 6.12: Interface for the Transmission Line Database 56