FITNESS ? Paving the way forward for GB?s digital substation journey

Authors: P. Mohapatra, C. Patterson and J. Mackenzie, SP Energy Networks, UK, C. Popescu, ABB, UK, and M. Wehinger, OMICRON electronics, Austria

The article gives insights on the Distribution Automation System (DAS) in Korea

In the modern society, electric utilities are facing many challenges in distribution systems like saturation of existing facilities, growing demand of the high reliability and high quality, increasing penetration of dispersed generations. Under this environment distribution systems are becoming more complex than before and its control and operation is becoming difficult using the traditional technologies and management methods. The remote control and operation of the power grid utilizing the communication and computer technology generates a lot of benefits for the grid operation and it has been widely used in the generation and transmission system level and this concept has been adopted in the low voltage level.

Distribution automation system (DAS) is a computer and communication system with necessary software and hardware facilities for remote monitoring and control of primary distribution feeders from a central server.

DAS collects system data measured from CTs, PTs and other field devices and monitors and assesses the system conditions and controls the grid. DAS continuously updates models of the distribution system in real time so grid operators can better understand distribution system conditions at all times. It acts not only as a decision support system to assist grid operators but also an autonomous agent in taking care of some urgent situations. The key results from deployment of DAS are higher service reliability, higher efficiency of operations and maintenance and reduced cost. Especially huge savings from deferred capital expenses can be obtained from optimized equipment use.

DAS can improve distribution system resilience to extreme weather events such as a storm by automatically limiting the extent of major outages and improving operator ability to diagnose and repair damaged equipment. It also provides a good platform for integration and management of distributed generations.

DAS has been in operation in Korea since 1997. Its development which started in 1991 has been evolving continuously and its history can be divided into four periods. The first period’s DAS started with a basic SCADA functions installed in a PC-based system. During this period, automated switches and RTUs were installed on the field. In the second period, DAS was expanded with many application functions and run in a fully equipped-control center. More advanced functions kept added in the third period and now DAS is going through the new era preparing for the smart grid era. As of today, owing to a long-time investment on DAS, 100 % of distribution feeders of the entire country are supervised and remotely controlled, and there are over 144 DAS control centers in Korea. The average restoration time is reduced to 6.2 minutes after DAS deployment from 73 minutes before. The description of DAS in this article is based on Korean DAS.

Components of DAS

Figure 1 shows the overall structure of the distribution automation system (DAS). There are four main components in DAS - Control Center, Remote Terminal Unit (RTU), Communication network and Automatic Switches. An effective distribution automation system combines complementary functions and capabilities and requires an architecture that is flexible or “open” considering integration with other system like EMS, OMS (Outage Management System), CMS(Customer Management System), AMR (Advanced Metering System) in a future.

Control Center:Figure 2 supervises the entire distribution system of concern. All of the decision making regarding the operation and management of distribution feeder systems are performed here. With the help of many decision-making aids, system operators try to maintain the desirable operating conditions. The most important job is to keep the power supply to the customers. The cost-effective operation of the whole system is another mission of operators at the normal situation.

The control center consists of DB servers with real time field data, facility data and history data, DA servers equipped with application programs, Front End Processor (FEP) that transfers analog and digital data obtained from the field to the DB servers and operator consoles where operators perform many operational activities via Graphic User Interface (GUI).

Remote Terminal Unit (RTU) is an intelligent device that has a physical connection to the field equipment such as reclosers and switches. RTU is responsible for properly converting raw data measured by sensors, to digital form for a modem to transmit the data and convert the received signals from the control center in order to control the equipment through actuators and switchboxes. It has the function of calculating rms voltage and current, real and reactive power, power factor, power quality measures, and a fault indicator, fault recording, etc. RTU for field device is usually called FRTU (Figures 3/4).

Automated Switches:  Conventionally many switches are installed on a feeder and some of them need to be automated so that their on/off status can be controlled remotely under a DAS environment. Automated switches are wired to FRTUs and their various operating status like on/off status, battery status, self-diagnosis status, etc. and voltage and current measurements at the installation point are sent to the control center by its associated FRTU via communications networks.

Open or close of automated switches changes the feeder configuration which is the main means in system operation. DAS server sends a switching command to FRTU, which then sends a signal to its associated automatic switch to actuate its operation (Figure 5).

Communication Network:  DAS operation relies heavily on a robust communication network to acquire large volumes of data FRTUs and to send control signals to FRTUs to operate field devices. Especially faulted location and service restoration which is the most important mission in DAS which requires fast, reliable and rugged communications. A variety of wired and wireless communication technologies to support DAS application have been tested and applied - PLC (Power Line Communication), telephone wire, optical fiber, pair cable, pilot wire, coaxial cable, radio, PCS, truncking radio, etc.

Many utilities prefer high-speed, fiber optic or microwave communications systems, while some choose to contract third-party telecommunication vendors for their high speed cellular network. Most common open protocols used for DAS are DNP 3.0, MODBUS, IEC60870-5-101. DAS in Korea mostly uses fiber optic communication with DNP 3.0 protocol (Figure 6).


Major Functions of DAS

Besides basic SCADA functions such as field data acquisition, monitoring of feeders and power equipment through GUI visualization, alarming, history logging, etc., many application functions for efficient system operation, control and management are performed in the DAS. There are two categories in these functions - base functions and application functions (Figure 7).

Base Functions

  • State Estimation: Accurate data are the key to effectiveness of application functions. Usually data contain errors that happen during sensing and communication. Especially voltage measurements on an overhead line have very poor accuracy because of the low precision PT installed in the pole. Thus, their actual values need to be estimated for the correct awareness of the system status and other applications.
  • Topology Processing:Identification of the current topology is another core element for almost all applications related to the system management. If any plan is set up with a wrong topology, associated switching could cause a fatal error in the system. Consistency checking of switch status with nearby measurements such as voltages and currents is commonly used for topology processing.
  • Power Flow Analysis:  This is the basic tool needed for operational planning which requires voltage calculation, current calculation, loss calculation, etc. When there is a need to change the configuration, voltage and current limit violation should be checked by power flow analysis.
  • Short Circuit Analysis: Since the operating conditions of the system are always changing, the protection capability of protection devices changes. Especially when the reconfiguration is performed in case of outage restoration, the fault analysis has to be performed for those devices in the area of reconfiguration and if necessary, they need to be reset.
  • Section Loading Estimation:Section loading data are the base information for any kind of reconfiguration applications. Protection setting and coordination need such information as well. Loadings in each section are estimated from measurements and maximum and minimum loading estimation is performed everyday using the last seven days’ loading data.

Application Functions:  The most important mission of the distribution system operation is to secure the service continuity of quality power to all customers. To be more specific, in the normal state, electric power of a good quality need to be delivered to all customers and in case of fault situation, the fault has to be removed immediately and resulting outage should be restored as quickly as possible. If any abnormal voltage or overloading happens in the feeder, the appropriate actions need to be taken in order to release the problems. If there is any significant change in the system topology or loading, the protection devices need to be checked for their protection performance and if needed, their settings need to be adjusted for maximum protection performance. DAS is equipped with such applications and some of the core applications are described below:

  • Fault Location, Isolation and Service Restoration (FLISR): When a fault happens, the protection device detects the fault and trips the circuit breaker causing the outage to the whole feeder or a part of the feeder. FLISR (Fault Location, Isolation and Service Restoration) is the process to handle the situation for full restoration of outage following a fault. Since the most important mission of the distribution system is to secure the power supply to customers, handling the after-fault situation is the most critical function in DAS and it is the core of the self-healing. There are two approaches - centralized and distributed, and they are described in details afterwards considering its importance.
  • Protection Coordination Analysis: Performance of the protection devices is analyzed based on the current or given settings of the protection devices. It can also determine the best settings for the given conditions. A screen below displays the information about the protection device and another one displays the coordination analysis (Figure 8).
  • Feeder Load Balancing:  When loads are evenly distributed among feeders, the integrated efficiency is obtained from the overall point of system operation. The feeder load balancing is done through moving loads of heavier-loaded feeders to other less-loaded feeders. So it is the process of finding the new configuration that achieves the maximum balance and one of the most commonly used method is the branch exchange scheme.
  • Loss minimization: Total loss of feeders could be reduced by redistributing loads among feeders. Loss minimization is the process to find a new configuration that achieves the minimum loss and can be done in a similar way to the feeder load balancing. Since the economy has become an important issue in the utilities, loss minimization function became attractive.
  • Fault locator: This is to pin-point the fault location after a fault is cleared. Precise fault location enables operators to dispatch repair crews to the right point and notify customers of outage status. The accurate fault locator has become an essential tool for improved operations and lower maintenance cost.
  • Crew Dispatch planning: When there is a need to dispatch crews for maintenance or restoration, integrating geographic information systems, traffic information and crew availability, repair crew can be dispatched in an efficient way using this function accelerating the works.
  • Switching sequencing: Once the plan is established for service restoration or maintenance work, the switching sequence of automated switches has to be determined in such a way that it does not cause any outage to the live area. 'open-before-close' rule is applied to the outage area for restoration and ‘close-before-open’ rule is applied to the non-outage area for transfer allowing a temporary closed loop.    
  • Volt/Var control: This is the optimization process taking into account all the optimization objectives in distribution analysis like minimum loss, number of control steps while keeping the voltage within limits. It could determine setting values of automated equipment like the tap of load tap changer (LTC), capacitor, voltage regulator and voltage amplitude of DGs.
  • Alarm Processing: Too many alarms generated in some cases do not help operators recognizing the events. Alarm processing is to transform the raw alarms into organized and timely information. Usually the minimal set of information is to be presented in a chronological way.

Self-healing Distribution System

A self-healing distribution system performs fault location, isolation, and service restoration autonomously without human intervention.

Although an intrinsic benefit of this technology is increased reliability, at the same time it is one of the key elements in ‘resilience’ and is expected to play a fundamental role in modern and future distribution systems. FLISR is the core of the self-healing system and it consists of a series of processes that are fault section identification, fault isolation and service restoration.

Centralized FLISR:  Fault indicator (FI) is mostly used to identify the faulted section. FI is the flag that is set when there is an overcurrent higher than a certain value and it is contained in FRTU as one of its functions. FI could be a separate device directly installed on the feeder.

A stand-alone fault indicator can be equipped with visual display to assist field crews and connected to communication network (Figure 9).

After a protection device trips the circuit detecting a fault, it informs the control center of a fault event. Then all FRTUs are requested to report their FI flags and a simple logic on topology using FIs can identify the faulted section (location). A command ‘open’ is issued to the switches surrounding this section for the fault isolation (Figure 10).

Service restoration starts with closing the feeder breaker (or recloser). For a load-side outage, the restoration plan needs to be established utilizing adjacent feeders. More than one possible restoration solution exists and the most suitable one can be determined by the fuzzy decision making or by operator’s selection. All of these decisions are made by the central master and commands are delivered to its slaves.

Distributed FLISR:  A more advanced self-healing system can be implemented using a multi-agent technology. Note that the previous restoration scheme adopts the centralized approach and its fault handling process is sequential, so completion of the whole process always takes some time. By applying the multi-agent system (MAS) technology, a distributed restoration can be performed resulting far reduced outage duration.

In this MAS-based distributed restoration scheme, every FRTU associated with switch becomes an intelligent agent (‘switch agent’) and the master station becomes ‘central agent’. Each agent has a capability of exchanging information with other agents.

It is provided a peer-to-peer communication and has its own intelligence and autonomous functions like self-recognition of fault section and fault isolation. The central agent has the role of providing necessary information and transmitting restoration strategies to the switch agents. In this system, any agent recognizing a fault makes the fault event public to all of the agents and every agent exchange FI information with its neighboring agents. Using their own intelligence on the collected FI information, they identify the faulted section and those switch agents surrounding the faulted section will issue an open switching command to its associated switches completing the fault isolation autonomously.

There are two approaches for the restoration planning - pre-planned approach and autonomous approach. Although the two approaches have a common scheme for fault isolation and switching execution, they are different in generating the restoration plan. The former utilizes the pre-planned plan and the latter develops its own restoration plan autonomously by cooperating with other agents. Although the second one implements the ideal multi-agent system concept, due to its unknown characteristics on the intermediate process and its final outcome, operators might feel uncomfortable and they may be reluctant to accept this approach in the early stage of MAS. The second approach has an advantage for operators’ acceptance.

In this approach for every fault case, the restoration plan is prepared in advance and its required switching actions are pre-planted into the associated switch agents. These plans can be generated by the service restoration planner and approved by operators before planting into switch agents. The central agent transmits a restoration plan and necessary information to the switch agents and each agent knows what to do when a specific fault happens.

When a fault happens, the switch agent associated with a protection device recognizes a fault and trips the circuit, informs the other agents of the fault after reclosing is finished, and those switch agents with FI of 1 (that means detection of a high current) exchange FI information with their neighboring agents.

By comparing FI information the agent can determine whether its associated section has a fault or not. Once the faulted section is identified, the agent will send out the fault section information to others and they start switching actions as pre-planted for restoration. All of these actions are performed autonomously (Figure 11).

Implementation of MAS-based FLISR:  DNP 3.0 used in Korean DAS has a master-slave structure and it does not provide P2P communication. In order to build the multi-agent environment, without replacing the entire communication network by one that can provide P2P, somehow one smart way has to be devised. DNP 3.0 DAS is added with a separate communication network (CDMA/PCS or Ethernet LAN) in a terminal agent called MASX. MASX is installed inside the FRTU and it keeps monitoring and analyzing the messages between the master server and FRTU  (Figure 13).

Recognizing the fault event, MASX starts into actions performing restoration functions as an autonomous agent. In this way, a FRTU with MASX becomes a true switch agent and the concept of the MAS can be implemented on the current operating DAS without affecting anything in its structure (Figure 12).

Testing of the MAS-based distributed restoration which was implemented into the existing DAS was performed in the Gochang KEPCO test center where integration tests, functional testing, performance evaluation before committing to DAS deployments are performed (Figure 14).

The total restoration time tested on a number of real faults generated by an artificial fault generator is recorded. An average restoration time of 48 seconds is marked for CDMA and 10.5 seconds for Ethernet-based wireless LAN, showing a remarkable reduction compared with the centralized scheme of 6.3 minutes.

In conclusion, owing to proven and anticipated benefits, DAS has quickly become a core process of a smart grid development around the world. There are some challenges for more extensive use of DAS like data integrity and synchronization, cybersecurity, interoperability, communication protocol.


Seung-Jae LEE (known as Paul) was born in Seoul, Korea and received his BS and MS in Electrical Engineering from Seoul National University, Seoul, Korea and his Ph.D. from the University of Washington, Seattle, USA.

He has been with Myongji University, Korea since 1988 and as a director of Next-generation Power Technology Center (NPTC). He was CIGRE SC B5 representative of Korea. He is an Editor in Chief of JICEE. He is the founder of APAP (International Conference on Advanced Power system Automation and Protection.) His main research fields are distribution automation, substation automation and protection relaying.

Let?s start with organization in protection testing