Comparison of the Operational Speed of Hard-wired and IEC 61850 Standard-based Implementations of a Reverse Blocking Protection Scheme

Abstract

This paper focuses on the reverse blocking busbar protection scheme with aim to improve the speed of its operation and at the same time to increase operational reliability, flexibility and stability of the protection during external and internal faults by implementation of the extended functionality provided by the IEC61850 standard-based protective Intelligent Electronic Devices (IEDs). The practical implementation of the scheme by the use of IEC 61850 standard communication protocol is investigated. The proposed scheme is designed for a radial type of a distribution network and is modeled and simulated in the DigSILENT software environment for various faults on the busbar and its outgoing feeders. A laboratory test bench is built using three ABB IEDs 670 series that are compliant with the IEC 61850 standard, CMC 356 Omicron test injection device, PC, MOXA switch, and a DC power supplier. Two types of the reverse blocking signals between the IEDs in the test bench are considered: hard wired and Ethernet communication by using IEC 61850 standard GOOSE messages. Comparative experimental study of the operational trip response speeds of the two implementations for various traffic conditions of the communication network shows that the performance of the protection scheme for the case of Ethernet IEC 61850 standard-based communication is better.

1. Introduction

An electric power system is divided into three major divisions which are known as generation, transmission and distribution. Electric power is generated and transmitted over vast territories which are exposed to all different types of conditions that might lead them to abnormalities and failure. These conditions can damage equipment and cause substantial loss of power. To avoid this problem each element in the system should be protected. The ideal function of the protective device is to notice the occurrence of faults and to isolate the faulted section from the power system as soon as possible in order to keep it stable [1].

Busbars are the most important components in the distribution networks. Faults on the busbar are un-common; however an occurrence of a busbar fault can lead to a major loss of power.

Busbars are the areas in a substation where the levels of current are high and therefore the protective relay application is very critical. In order for the protection scheme to be successful it is important to carry out the following specifications: Selectivity, Stability, Sensitivity, and Speed. To meet all of the above requirements protection must be reliable, meaning that the protection scheme must trip when called up to do so (dependability) and it must not trip when it’s not supposed to (security) [2-4].

Traditionally protection and control devices used hardwire as a path to distribute signals between each other. This method of communication created time delay response from the sending device to the receiving devices. The delay came from the time taken by the auxiliary relays for switching on and off when the signal had to be sent from a binary output of the sending one into the binary input of the other device. By means of the new digital technology these traditional reverse blocking schemes can be considerably improved in terms of operational speed. The new digital protection also offers an increase of flexibility and operational reliability. This paper focuses on improving the operation of reverse blocking scheme used in distribution busbar protection. This is achieved by introducing Generic Object Orientated Substation Event (GOOSE) messages between the IEDs that are interconnected by a Local Area Network (LAN). This network allows the blocking signals to be sent directly from an IED to another IED without any additional delay created by the auxiliary relays [5-8].

In order to prove the above statement a distribution busbar reverse blocking scheme is designed, simulated, and implemented. A DIgSilent software package is the suitable program that is utilized for the simulation. A test bench is designed and implemented to investigate the real-time operation of the designed protection scheme. Simulation and real-time implementation investigations of the operation of the reverse blocking scheme are done using IEC 61850 standard compliant IED models or devices.

Two case studies are considered during the simulation and real-time operation:

Case study 1: The communication between the IEDs is done through hardwires. Case study 2: The communication between the IEDs is done through Ethernet GOOSE messages

A model of the distribution network is selected and its parameters are calculated. The fault currents are calculated for different fault locations. Various types of faults are applied to the network in order to be able to configure the setting of the protective devices. ABB relay models from the DIgSilent library and ABB IEC 61850 standard-based IEDs are used respectively for simulations and real time implementation of the protection scheme. The effect of the different fault levels over the fault clearing times and the stability of the network are investigated. The protection devices time grading and co-ordination settings are analyzed in order to have an efficient busbar protection scheme. The results are compared and analyzed.

The content of the paper is presented as follows: The model of the distribution network is introduced in part 2. Calculation of the short circuit currents is done in part 3. Short circuit simulation is presented in part 4. IEC 61850 standard requirements to the digital relays (IEDs) are presented in part 5. Part 6 presents the IEC 61850 standard based operation of the busbar protection scheme. The development of the laboratory test bench is described in part 7. Part 8 presents the implementation of the practical experiment. Communication establishment between PCM 600 software and the IED’s is described in part 9. Part 10 presents the GOOSE engineering of the laboratory test bench. Part 11 presents the practical experiment result for the fault at the outgoing feeder. Discussion of the obtained results is done in part 12, and part 13 presents the conclusion.

 

2. Distribution Network Modeling

A distribution network is constructed and modeled with its relevant equipment and it is simulated. This network has generation, buses, and a transformer, network connections, cable interconnection, and loads. Load flow studies are carried out for the distribution network. The basic requirement for a distribution network is to determine the total fault level and the contribution of the short circuits within it. In a radial network generally for medium voltage (MV) and low voltage (LV) the fault current contribution is determined by the short circuit impedance of the transformer. When a short circuit occurs within the network a high fault current is created that leads the network to fail. Therefore protection devices are needed to protect the distribution system not to fail when faults arise. Overall these elements influence the power quality performance of the network. The distribution network that is studied is shown in Fig. 1 with its data in the form of One Line Diagram (OLD). The study of the network is focusing on the fundamentals of single phase and three phase short circuit currents calculation and how these currents contribute to the network protection.

Fig. 1.One line diagram of the studied distribution network

2.1 Simulation of the load flow in DIgSILENT software

Planning, control and operation of the distribution network require load flow calculations in order to analyze the steady state performance of the system. This is implemented under various operating conditions and equipment configurations. The objectives for the simulations are to develop different case studies for load flow analysis, and short circuit analysis in order to design the protection of the distribution network system. This load flow calculation determines voltage magnitude V, voltage angle ϑ of the nodes, and the active P and reactive Q power flows on all branches. An AC balanced, positive sequence load flow is performed making use of the Newton-Raphson method for current and power equations simulation. Parameters used in the simulation are shown in Table 1.

Table 1.The parameters of the considered distribution network

The network system model in DIgSILENT is shown in Fig. 2 where the result boxes in the blue color represent the active, reactive and apparent power below. The result boxes in the red color represent voltage in kV, voltage in p u and the phase angle.

Fig. 2.One line diagram of the load flow calculation for the studied distribution network

2.2 Load flow results

The load flow results for the three phase voltages and three phase currents are shown in Fig. 3 below where phase A is represented in red, phase B in yellow and phase C in blue. The results show that the voltages and currents are balanced with a 120 degrees phase shift. This is recorded during normal condition in the distribution network. Table 2 shows different voltage levels in the network from busbar 1 up to busbar 4, Table 3 shows voltage deviation profiles of the calculated load flow which are within the IEEE standard 141-1993 voltage criteria [9].

Fig. 3.Three phase voltages and currents for the simulated distribution network

Table 2.The load flow results for the different voltage levels in the network

Table 3.The load flow results for the busbar voltages

The results analyses show that voltages at the load buses are generally averaging around 0.96 per unit. The loading on a transformer and cables are within normal working capabilities. The generator reactive outputs are within the limits that are defined by the generator capability curves. The voltage criteria are within the IEEE standard 141-1993 requirements [9].

The next step is to calculate the short circuit currents as a preliminary step before the design of the protection scheme.

 

3. Calculation and Simulation of the Short Circuit Currents

The load flow study is used to in order to be able to specify fault ratings for all electrical devices that are used for protection. Therefore short circuit current calculations in the distribution network are implemented in order to identify other potential problems. There are five general steps that are followed when calculating short circuit currents based on the IEC 60909 standard in the distribution network. These steps are as follows [10]:

Construction of the distribution network and its relevant parameters (done on previous part) Calculation of the short circuit impedance for all relevant components Reference of all impedances to the reference voltage Calculation of balanced three-phase short circuit currents Calculation of single-phase to earth short circuit currents Detailed equations and results from the calculations can be found in [16]

3.1 Short circuit simulation

When protection is designed and implemented in the network system it is essential to know the fault current distribution throughout the system. This is obtained by simulations. The purpose of these simulations is to be able to specify fault ratings for all electrical devices that are used for protection. It is important to know voltages due to the fault in the different nodes within the network. It is also required to know the current amplitude at any relaying point so that the fault is to be cleared with precise discrimination. The parameters that are required for each set of faults are as follows [11]: Where the fault is applied, which phase is affected, which feeders are affected, which elements are affected, current and voltage sequence, maximum fault current, and minimum fault current

In this case the interest is on the maximum expected currents and on the minimum expected currents to manage the design of the protection scheme. The tool used for the short circuit calculation and analysis is the DIgSILENT simulation software. Different case studies are developed with utilization of a Single-Line-Diagram (SLD) tool. The short circuit location and its requirements are chosen. The initial three phase short circuit current , single phase to earth short circuit current , and the peak short circuit current Ip of the system operation are simulated and taken into consideration. The maximum and minimum short circuit current values are calculated using the IEC 60909 standard. The short circuit current calculation for the fault levels involves also the short circuit power . The equations used are represented as follows [10]:

where is the initial symmetrical short current, IP is the peak short circuit current, k is the constant factor, c is the voltage factor , Vn is the nominal voltage at the short circuit location Zk is the equivalent short circuit impedance, Zk(1) is the equivalent positive sequence short circuit impedance, Zk(2) is the equivalent negative sequence short circuit impedance, Zk(0) is the equivalent zero sequence short circuit impedance

The short circuit currents are applied in to the distribution network at specific locations namely: busbar B2 (substation2/2), load 1(line B2 to B3) and load 2 (Line B2 to B4) respectively as shown in Fig. 4 where boxes in the red colour represent the short circuit power, the initial symmetrical short current and the peak short circuit current.

Fig. 4.One line diagram of the studied faults in the distribution network

The three phase network is normally treated as a balanced symmetrical system. When the fault occurs in the power network systems the symmetry is normally disturbed as shown in Fig. 5 below. This situation results in unbalances in the current and voltage within the network. The only case that is exceptional is when a three phase fault occurs, because all three phases are involved. Investigation of the influence of every one of the considered faults is described below.

Fig. 5.Three phase voltages when short circuits are introduced in the distribution network into a busbar B2 (substation2/2), a load 1(line B2 to B3), and a load 2 (Line B2 to B4) respectively

Table 4 represents the simulation results for all fault types and locations in the distribution network. The results give an indication for all the short circuit currents at their locations where the protection devices have to be installed. From the results it is now possible to select a specified setting and fault rating of the protection devices that are used for the scheme.

Table 4.The short circuit simulation results

 

4. Selection of the Busbar Protection Scheme Characteristics and its Operation Investigation by Simulation

The overcurrent setting for the protection devices is done by selection of some parameters. These parameters require time-current characteristics for both instantaneous and time delay elements. The relay characteristics generally should all confirm with BS142/IEC or ANSI/IEEE standards. Simulations are performed according to the IEC60909 standard.

A busbar reverse blocking scheme is investigated in which time discrimination between the incoming feeder and the outgoing feeder protection is sustained. This method required that all relays must be equipped and set with time-overcurrent, instantaneous overcurrent, time-overcurrent earth, and instantaneous overcurrent-earth Inverse Definite Minimum Time (IDMT) and Definite Time (DT) elements. The incoming feeder curve is required to be slower than the outgoing feeder curves for all fault currents. It is also important to have a suitable grading time margin between the two curves with 0.4 seconds for the electromechanical relays and 0.3 seconds for the digital relays [12-14]

The designed settings of the protection devices from the reverse blocking scheme are used to investigate the quality of the protection operation in the DigSILENT software environment for the considered distribution network. Two case studies are considered on the basic of the ABB protection devices. Case study 1: Models of conventional relays. Case study 2: Models of IEDs compliant with the IEC61850 standard (with Ethernet communication) Simulations of the operation of the reverse blocking scheme are done for every one of the two cases. Organisation and running of the simulations is the same for the both cases as follows:

One relay is installed at the main incomer on the secondary side of the transformer, named (Relay C). One relay named (Relay A) is installed on the outgoing feeder that feeds load 2 and the other relay named (Relay B) is installed on the outgoing feeder that feeds load 1. The aim of this study is to execute a short circuit current in the network where each relay is located making use of the IEC 60909 standard. Two sets of simulations are done for each protection device, namely minimum short circuit with single phase to ground fault type and maximum short circuit with three phase fault type. A combination of inverse time and definite time characteristics are used for the busbar protection. The inverse time is selected for longer grading time. The define time element is selected for the shorter grading time. The reason for this is to have a protection scheme that is effective and has high sensitivity.

The three phase fault in the line (B2 to B4) known as external fault for the busbar B2 protection scheme is simulated as shown in Fig. 6 (a). The data shown in the result boxes in the blue colour represent the short circuit power, initial symmetrical short current and peak short circuit current. Fig. 6 (b) shows the recorded results for the three phase voltages and currents in the line (B2 to B4) when a short circuits are introduced in the distribution network. The execution time for the short circuit event is 0.1 seconds. During the short circuit period the phase current increases and the phase voltage decreases creating the system to be unbalanced. The maximum fault level of 315.97MVA, a short circuit current of 16,584kA and a peak current IP of 44,089kA are recorded. After the 0.1 seconds the fault is cleared in the line (B2 to B4) and the phase voltages and currents go back to their normal condition

Fig. 6(a).Three phase fault between a Substation2/B2 and a Substation4/B4

Fig. 6(b).Three phase voltages and currents in the line (B2 to B4) when three phase fault is introduced in the distribution network

A time/current curve of the Relay A is used to analyse the results and the performance of the protection scheme. The two types of elements were used for this curve namely Inverse Definite Minimum Time (IDMT) and Definite Time (DT) show for the three phase fault simulated in Fig. 6 (a) that the tripping times obtained for the relay A are 0.100 seconds for the DT and 0.324 seconds for the IDMT elements as shown in Fig. 7 above. The earth element displays 9999.99 seconds for the tripping time. This fault only affects the phase fault currents therefore only the phase faults elements tripped

Fig. 7.Relay A protection response for the three phase fault with a clearing time of 0.100 seconds for the DT and 0.324 seconds for the IDMT element respectively

Fig. 8 above shows the simulated tripping characteristics for both feeders line (B2 to B3) for the load 1 and the line (B2 to B4) for the load 2 when a three phase fault occurs. It can be seen from the results that the relay B shown in green does not see the three phase fault that occurs in the line (B2 to B4). Therefore Relay B tripping time is displayed as 9999.999 seconds for both the IDMT and the DT elements. This means that for the fault in line (B2 to B4) the relay B did not trip. This ensures that the protection is correct and no false tripping or unwanted tripping will be experienced when the scheme is implemented.

Fig. 8.Protection responses of the Relay A and B to the three phase fault

The three phase fault indicated in Fig. 6 (a) is picked-up by A and C protection devices. With utilization of the IDMT phase element for the relays A and C, a sequence tripping response of 0.324 and 0.810 seconds is recorded as shown in Fig. 9. The earth elements tripping times are 9999.99 seconds, meaning that they did not trip for this fault. The results also show that the relay A which is close to the fault responses in 0.324 seconds while the direct upstream relay C follows with a response delayed by 0.810 seconds. This algorithm suggests that should the relay A fail to operate then the next up-stream relay C will operate in 0.810 seconds. Therefore the relay C is used as a back-up protection for the relay A. For faster tripping time the DT characteristic of the relay A responds in 0.100 seconds and the instantaneous characteristic for the relay C delays by 0.420 seconds.

Fig. 9.Characteristic of the Relay A time-overcurrent response and of the Relay C acting as a back-up protection device for the three phase fault

The results for the considered two cases for all different faults and their locations for the conventional relays are all presented in Table 5 as they are exactly the same. The above experiments are repeated with the models of the IEC 61850 based IEDs and the same results are obtained. The reason for this is that the simulation environment of DIgSILENT uses the same models of the protection elements for both conventional and the IEC 61850 standard-based protective devices and the communication between the relays is done by the same software

Table 5.Combined results for the different faults in the distribution network using conventional relays

Implementation of the protective scheme using real protective and a conventional or IEC 61850 standard-based approaches for exchange of data signals between the relays is expected to bring differences in the relays time responses

 

5. IEC 61850 Standard Requirements to the Digital Relays (IEDs)

In the Substation Automation System (SAS) all protection related activities are influenced or changed by introduction of the IEC61850 standard. The main idea is for the user to understand how to implement the standard in practice [7]. It is also important to have technology that is cost effective when designing SAS to ensure faster returns on investment. The SAS must comprise of devices that can supply a wide range of communication capabilities and interfaces compliant to IEC 61850 standard. It also requires devices that ensure maximum performance in protection, control and monitoring. The ABB’s new IED 670 series is capable of meeting the above requirements [15] and that is why it is selected for the practical implementation of the designed protection scheme

The ABB IED 670 has adopted new algorithm for numerical calculation module. The new alternative Central Processing Unit (CPU) model design consists of the main Controller/CPU type IBM 3200, power computer 750FX 600MHz, and internal 100 Mbit/s communication bus components. The memory capacity is 128 Mb FLASH. It permits performance of one millisecond for differential protection functions, three milliseconds for other protection functions, and for the logic operations it takes between three to hundred milliseconds depending on the application [15].

The IEC 61850 standard communications for ABB IED 670 is established on Ethernet network with a speed of ten to hundred megabits per second. When the IED 670 series processes information it utilizes IEC61850-8-1 part of the standard for low and medium speed and IEC61850-9-2 part for high speed. The IEC communication set up is designed in a network system and each IED must have a specific internet protocol (IP) address [15]. The communication network is used to transport data into different environments within the substation. The substation consists of a number of components at the different levels. Through the network these levels are required to communicate with each other, as shown in Fig. 10 below [8]. The levels are: process (where data acquisition, sensors and actuators are found); bay (where protection, monitoring and control task are found), and station where the SCADA operations are performed. The communication that occurs between the levels makes use of IEC61850 standard [15].

Fig. 10.Substation Automation system architecture

 

6. IEC 61850 Standard Based Operation of the Busbar Protection Scheme

For this protection system the incoming feeder IED and the outgoing feeder IEDs communicate using IEC 61850 GOOSE messages to operate the protection and control functions. When a fault occurs at one of the outgoing feeders a multifunctional IED will trip the feeder breaker and send a GOOSE message to block at once all the incoming IEDs that are not involved. The operational reliability of the protection scheme is based on GOOSE messaging. A laboratory test bench had to be developed as a platform where the practical experiments are exercised.

 

7. The Development of the Laboratory Test Bench

A laboratory test bench is developed in order to be able to implement the designed busbar protection scheme known as a reverse blocking scheme. The IEC 61850 standard is implemented on the scheme and it is investigated how it influences the busbar protection scheme performance The Laboratory test bench consists of the following components: ABB RED 670 IED (IED_C), ABB REL 670 IED (IED_B), ABB REB 670 IED (IED_A), PCM 600 configuration software tool, CCT600 communication software tool, CMC356 test injection device, MOXA switch, DELL computer, and auxiliary power supply.

All experiments in this paper are developed using this test bench. The main aim of the experiments is to test and compare the communication performance between the IED’s using both hard-wired systems for transfer of data signals as shown in Fig. 11a and GOOSE messages based on the IEC61850 principles, as shown in Fig. 11b.

Fig. 11(a).A busbar reverse blocking scheme using a hard-wired signals

Fig. 11(b).A busbar reverse blocking scheme using IEC 61850 standard-based communication

 

8. Implementation of the Practical Experiment

Fig. 12 shows a modified compact flow chart that is used for the practical experiments implementation. The IED configuration is done using PCM600 software, the tool for operation and configuration of all the ABB 670 series protection devices (IED’s). This software can be used to parameterize the device, view process data and evaluate fault that is recorded. With this type of service the information for different events can be speedily transmitted between IEDs

Fig. 12The flow chart for the practical experiment implementation

Fig. 12 shows that communication of IEDs needs to be first configured. The next step is to configure the IED application using various function blocks of the PCM600 tool Implementation of the steps of the flow chart is described in the points 9 and 10 below.

 

9. Communication Establishment Between PCM600 Software and the IED’s

The used communication protocol within the substation is dependent on the communication between the IED and the PCM600 software. The IEDs have an Ethernet interface port situated on the front panel which is used for communication with PCM600 software. These IEDs also have Fibre Optic ports situated at the rear panel but the developed lab test bench only focuses on utilizing the Ethernet interface port. When it is used for communication the TCP/IP protocol is employed in order to implement the protection scheme.

9.1 The computer set up to access multiple Intelligent Electronic Devices (IEDs) via a switch

This setting algorithm is implemented when developing the lab test bench. It makes use of an indirect connection through a station LAN. Once the physical connection between the IEDs and the switch is completed, a physical link via a switch and the PC has to be added as well. Then the network connection from the PC had to be configured using the same procedure for the direct point to point connection. Since the three IEDs are now linked to the PC via a MOXA switch as shown in Fig. 13 their IP addresses are changed and the corresponding sub-network mask is not necessary to be changed.

Fig. 13.Setting up the PC to access more than one IED via a MOXA switch

 

10. GOOSE Engineering of the Laboratory Test Bench

Once the IEDs configuration is completed in the PCM 600 software the next step is to export the project at the substation level of the plant structure as SCD file. This file contains all communication and substation configuration of the IEDs. This file is then imported to the CCT600 software tool to create the IEC 61850 GOOSE messages communication. This software tool is utilized to configure GOOSE communication and is used as a platform for mapping of all the data that is needed for the project. When the GOOSE engineering is completed in the CCT600 software it is again required to export the SCD file from the CCT600 software and to import it back to the PCM600 software. When this is completed GOOSE connection between the used IEDs is required to be done at the Signal Matrix Tool that is found in the PCM600 software. Once this procedure is completed the configuration is written to the three IEDs. The algorithm for this procedure is shown in Fig. 14 below. The steps of this flow chart are described in points 10.1 and 10.2

Fig. 14.The flow chart algorithm for GOOSE engineering of the laboratory test bench

10.1 GOOSE engineering using CCT600 software tool

After the SCD file is exported from the PCM600 software it is then imported into the CCT600 software. Once the SCD file is imported to the CCT600 software tool the IEDs can be viewed at the project navigator window. At the communication section the IED_C is represented with a technical key AA1J1QO1A1, the IED_B is represented with a technical key AA1QO2A1 and the IED_A is represented with a technical key AA1QO3A1.Once all IEDs are identified the next step is to create the data sets under the IEC61850 Data Engineering window that are going to be sent as GOOSE messages by the IEDs.

The next step is to organize the Logical Devices, Logical Nodes, Data Objects and Data Attributes for the IEDs that are used. This communication section is found under Logical Device (LD) at the Logical Node Zero (LNLLNO). The LN LLN0 is a special LN per LD that contains the DataSets and the various control block inputs coming from the GOOSE messages. Once the data sets are configured the next step is to do GOOSE control engineering. In order to be able to send GOOSE messages from the IED_A and IED_B to the IED_C, the GOOSE Control Block (GoCB) is defined for the IED_A and IED_B. This GoCB contains Datasets that carry data objects of data attributes shown in the Fig. 15. When the data attribute triggers changes as a result of a fault, GOOSE messages will be published and IED_C at the incoming feeder is subscribed to them. Therefore all the data sets that are configured for the IED_A will be replicated into the buffer that is going to be sent. The buffer will also contain the actual value and will then be sent as a message. The subscriber which is IED_C will then receive the GOOSE messages including their sequence numbers. Once this configuration of the GOOSE control engineering is completed routing of Information between the IEDs and the GoCB is done.

Fig. 15.The mapping of GOOSE control engineering

The IED_C which should receive the GOOSE message is recognized. This IED has to be described in the engineering state as one that will receive GOOSE messages. It is defined at the LN0 under the structure of the LD in the receiving IED_C that the sending IED is IED_B. This IED is identified as the input. This link is done by drag-and-drop method of the IED icon to the GoCB shown in Fig. 11. When the linking of GoCB to an IED is completed the next step is to update the dataflow and the output results.

10.2 Signal Matrix Tool (SMT) in PCM600 software used for making GOOSE connection

The SCD file from the CCT600 software is imported back to the PCM600 software tool in order to make GOOSE connection. This configuration is done using SMT. When the SCD file is imported to the PCM600 software under SMT a new page is automatically created and it is called GOOSE receive. In this case the page shown in Fig. 16 represents the GOOSE connection between the IED_C and the IED_B. The output of Four Step Residual Overcurrent Protection (EF4PTOC) function block (START and TRIP) and also Four Step Phase Overcurrent Protection (OC4PTOC) function block (START and TRIP) are sent via GOOSE messages to the IED_C. The same method is used for connection between the IED_C and the IED_A. The configuration is completed and the next step is to send the data to the IEDs.

Fig. 16.The representation of the signal matrix tool in PCM600 software

The two configurations: IED configuration using PCM600 software tool and IED GOOSE engineering using CCT600 software tool are now implemented. The next step is to run the reverse busbar blocking scheme using the laboratory test bench and to record the results.

 

11. Practical Experiment Results for the Fault at the Outgoing Feeder

11.1 Logic of operation of the reverse blocking scheme

For this lab-scale implementation a three phase fault current of 15717.25A is injected to the outgoing feeder where IED_B is installed. The IED_B using the DMT characteristics parameters starts and sends block signal to the incoming IED_C on the incoming feeder with a delay of 0.2 seconds for a TRIP signal to open the breaker that is sitting in the outgoing feeder shown in Fig. 17 below. The IED_C sitting at the incomer bay is blocked until the operation of the IED_B is completed. In a case where the IED_B does not trip the IED_C will trip the breaker that is situated on the incomer bay.

Fig. 17.The three phase fault at the outgoing feeder where the IED_B is installed

11.2 GOOSE communication vs Hard-Wired (HW) connection

There are two types of tests implemented. One is that the IED_C is linked with the IED_B and the IED_A via hard wire (the normal traditional method) and the second one where IED_C is linked with the IED_B and the IED_A via an Ethernet cable using a GOOSE based method where GOOSE messages are used to block the fast overcurrent element of the IED_C from operating. The objective for this task is to measure how fast the IED_C protection overcurrent element can be blocked in the event where a fault occurs on the outgoing feeders. In order to achieve this objective a minimum operating time for the high-set overcurrent protection for the IED_C needs to be set properly. The results shown in Fig. 18 is the time taken by the blocking signal that is sent by the IED_B sitting on the outgoing feeder to the IED_C sitting on the incomer feeder. From the moment the IED_B is triggered the GOOSE BLOCK signal represented in blue in Fig. 18 is sent to the IED_C with no time delay and the hard wired BLOCK signal, represented in red, is delayed by 24.2 ms as shown below.

Fig. 18.Comparison of the time for blocking of the overcurrent element of IED_C for the hardwired and Ethernet based communication

The results recorded when both blocking signals were already sent to IED_C are shown in Fig. 19. In this case the GOOSE BLOCK represented in blue shows the binary signal measured to see how long it took before the TRIP signal is issued by IED_B at the outgoing feeder. It can be seen that the GOOSE message took 216 ms delay from time when IED_B was triggered which is correct based on the DMT characteristic that was selected for the IED_B.

Fig. 19.The results for the time measurement taken by the GOOSE BLOCKING binary signal

In Fig. 20 the HW BLOCK represented in red shows the binary signal measured to see how long it took before the trip signal is issued by the IED_B at the outgoing feeder. The result shows that the HW BLOCK signal took 240.8 ms time delay and based on the DMT characteristic set for the IED_B the time delay is increased by 40 ms. In a case where the fault is at the feeder where the IED_A is installed the same procedure is implemented.

Fig. 20.The results for the time measurement taken by the HW BLOCKING binary signal

11.3 Description of the tests that were done

A typical distribution network with conventional hardwired (HW) blocking signal path connected between the three protection IEDs shown in Fig. 11a is used for this test. Two different types of faults namely phase to ground and three phases are each injected at the feeder where IED_B is installed. An Omicron CMC 356 simulator is used to inject the fault currents. The next task is to replace the conventional HW blocking signal path with Ethernet LAN to use the GOOSE service of the IEC 61850 standard as the signal path for the busbar protection scheme as shown in Fig. 11b. A star network topology have been utilized for scheme implementation. Two different types of faults are again injected. The experiments are done in order to measure how fast the incomer protection IED_C can be blocked in the event when the fault is at the outgoing feeder where IED_B is installed. The experiment are repeated 10 times they show the same result of 1344.86ms for the time taken when using conventional HW blocking signal path and 1320.16ms for time taken when using GOOSE messages (Ethernet LAN) as the blocking signal for all 10 times repeated tests. The average values (meaning sum of the 10 results obtain was divided by 10 in order to get the average values) of the results are analyzed and recorded in Table 6 where the time delays between the conventional HW blocking signal path and GOOSE messages (Ethernet LAN) blocking signal are consistent for both phase to ground and three phases faults.

Table 6.Results for the average time taken by the reverse blocking signal for different faults in the distribution network

11.4 Operational speed test under communication network burden

A typical distribution network with conventional hard-wired (HW) blocking signal path connected between the three protection IEDs shown in Fig. 11a is used for this test. Two different types of faults namely phase to ground and three phases are each injected separately at the feeder where IED_B is installed.

Three more IEDs from SIEMENS are connected into the network through the same switch. One IED is configured to publish GOOSE messages which are broadcasted periodically in every one second just to create more traffic on the network. The other two are configured to subscribe to the GOOSE messages. The network now has six IEDs that are connected together through a switch and the experiment explained in section 11.3 is repeated. The loading of the three IEDs is done to see if the speed performance of the reverse busbar protection scheme in real-time is affected. A network packet analyser known as Wireshark is used to capture GOOSE messages in realtime and display those in a readable format [17-18]. It is also used to see the background traffic and all other configuration parameters in real-time. The results in Table 7 show that with additional data traffic in the network the speed performance of the reverse busbar scheme is not affected for the considered case. This conclusion is done based on the fact that the results in Table 7 are the same as the results in table 6, where there no additional data traffic in the network. Table 7 shows the time taken when using conventional HW blocking signal for single phase to ground fault to be 1344.86ms and 240.8ms for three phase fault. This means that additional data traffic does not affect this path at all. For this particular case study the time taken when using GOOSE messages (Ethernet LAN) is 1320.18ms for the single phase to ground fault and 216.4ms for three phase fault. The time delay between the conventional HW blocking signal path and GOOSE messages (Ethernet LAN) blocking signal is also at least 24ms which is same as the previous results shown in Table 6.

Table 7.Results for the time taken by the reverse blocking signal for different faults in the distribution network with additional data traffic

This project focused on the test bench platform for one station where one internet switch was used with six IEDs in the same broadcast domain. All the devices were housed in one rack and therefore it was not necessary to apply VLAN configuration because there was no need to transport data in multiple VLANs.

A LanTraffic V2 software tool is then used to generate more traffic using the following protocols: Transmission Control Protocol (TCP) and User Datagram Protocol (UDP). A testing configuration was made using two computes for implementation where computer A is used for sending traffic and computer B receives the traffic. These computers both have installed LanTraffic V2 software tool and are linked via the same switch that is used for the protection scheme as shown in Fig. 21 below. Three types of methods are implemented for generating traffic namely: automatic data generator by using mathematical laws, packet generator where different parameters are defined and sent, and file method where a selection of a file is sent.

Fig. 21.The computer A and B both installed with LanTraffic V2 software tool and are linked via the same switch that is used for the protection scheme.

Four steps were followed when the LanTraffic V2 software tool was launched. Step one was to identify how the traffic was going to be generated. There are two options that can be selected namely: using one computer (Sender and Receiver) or using two computers where one generates the traffic (Sender) and the other one receives the traffic (Receiver). In this case two computers were used where one generates the traffic and the other one receives the traffic. The second step was to configure the LanTraffic V2 using two connection protocols were first one is the TCP protocol and the second one is the UDP protocol. The configuration is based on setting up the IPv4 destination address which in this case it is 10.1.150.8 and it port number is 8080 for the sender. The third step was to set the port at the receiving computer to be the same as the sending computer. The fourth step was to select the types of methods that are implemented for generating the required traffic as shown in Fig. 22.

Fig. 22.The traffic generator parameters in the LanTraffic V2 software tool

Once all the above was completed the traffic was generated from computer A via the switch to computer B. While the traffic was being generated the results were also recorded in real time and Fig. 23 shows the throughput, volume and packets of traffic data from the computer A (sender) and Fig. 24 shows the throughput, volume and packets of traffic data from the computer B (receiver).

Fig. 23.The throughput, volume and packets of traffic data from the computer A (sender)

Fig. 24.The throughput, volume and packets of traffic data from the computer B (receiver)

During this process the reverse blocking scheme was implemented repeatedly while the traffic was generated into the network. The traffic generated was up to volume of 17.1 GB and 13139 packets but the operation speed of the scheme was consistent and not changing because GOOSE messages were tagged as the high priority ones based on IEEE 802.1Q when compared to other traffic that was published into the network.

 

12. Discussion of the Obtained Results

A comparison between GOOSE communication and the HW connection is performed. The results show that when using GOOSE services the response time from IED_B to IED_C is considerably faster than traditional hard wired one. This fact is due to the response time of the hard wired to be constrained by the auxiliary relay on/off delay of the sending IED_B and also by the input filtering of the receiving IED_C. These delays are omitted when using GOOSE communication. The results show that the operation speed of the schemes is gained when using Ethernet LAN for the implementation of the blocking signal path. The feasibility of the Ethernet network has been analysed with additional data traffic and it was proven that it has not affected the speed performance of the busbar reverse blocking scheme.

In addition the reliability characteristics of the protection scheme are increased on the basis of the organisation of the GOOSE communication. GOOSE messages are broadcasted periodically even when no data change occurs. This periodic broadcasting enables the GOOSE communication to be constantly monitored. When single or a number of the data attributes within the GOOSE dataset change, the first broadcast with the updated data values is sent. After the first broadcast the same GOOSE message is repetitively broadcasted into the network. The advantage of using GOOSE messages is that the receiver IED can sense communication losses if the periodically sent message disappears. This concept enables GOOSE communication to be supervised, which is not possible with hardwired solutions. The disadvantage of using HW is that when the number of IEDs increases the number of wires that is used for information exchange between the IEDs increases excessively. The Advantage of using GOOSE communication instead of conventional hardwiring is that it decreases the number of cables used for wiring. Using GOOSE messages for implementation of the scheme makes it easily to be extendable because it is implemented through software configurations rather than hardwire signal paths.

There were few factors that introduce constrains in the simulations. The DigSILENT software has built in circuit breakers and therefore it was simpler to get the accurate operating time for the full complete scheme. On the other side in the DigSILENT software based simulation the GOOSE communication could not be implemented. The communication between devices via GOOSE was only done on the practical implementation of the busbar protection scheme. The communication system between the protection devices is successfully configured and desired results are obtained. The traditional protection scheme for the busbar where hardwire is used to link the protection devices is investigated and the three factors that create constrains and time delays in the scheme operation are evaluated. The first factor is the delay made by the auxiliary power supply energising the path for signalling. This delays it typically 24 ms in this case. The second delay is the time taken for filtering and handling the input signal coming from another IED. Usually when the power is connected to the contactors an electromagnetic disturbance creates a spark and therefore an additional filter is needed and this creates another time delay. When the GOOSE messages are used for horizontal communication all these problems are solved. For a signal to be sent from one protection device to another one through the binary modules of the IEDs there is no need of auxiliary power supply and therefore no delay is created. This way of communication also simplifies the wiring and signalling of the system. The lab test bench developed is able to improve the speed and reliability of the protection scheme making use of GOOSE messages for communication.

 

13. Conclusion

The effectiveness of using the IEC61850 standard for protection is investigated and analysed. The IEDs configuration is successfully achieved with an assistance of the PCM600 software tool. The GOOSE engineering for the IEDs was also achieved successfully. The new approach of replacing the traditional conventional hard-wired blocking signals between the protection devices by an Ethernet LAN for GOOSE communication was proven to be successful. The operational speed was improved by using the GOOSE messages, when compared to the operational speed of the hardwired blocking signal for the busbar protection scheme. The reliability of the busbar protection system is also improved. The new method furthermore creates flexibility for future improvement of the protection schemes. It means that it provides the system to be easily extended and also reconfigurable for different network topologies. The test bench will be utilized for training of students and employees from utility companies that are in the protection field. It can be used as the platform to perform different schemes and to test the performance of GOOSE communication between the IEDs.