Home / Technical Articles / Communications in power system protection (medias, protocols and network topologies)

Communication system

A communication system consists of a transmitter, a receiver and communication channels. Type of medias and network topologies in communications provide different opportunities to advance the speed, security, dependability, and sensitivity of protection relays.

Communications in power system protection - Media, topology and protocols
Communications in power system protection - Media, topology and protocols (on photo: 110kV-20kV substation protection cabinet; credit: Marko Gostovic via Linkedin)

There are a several types of communication media such as micro wave, radio system, fiber optic, etc. The advantages and disadvantages in communication medias which are currently in operation (both analog and digital) and different network topologies are summarized below, respectively.


  1. Comparison of Communication Medias
  2. Comparison of Different Communication Network Topologies
    1. Point-to-Point network
    2. Star network
    3. Bus network
    4. Linear Drop and Insert network
    5. SONET Path Switched Ring
    6. SONET Line Switched Ring
  3. Description of Different Communication Protocols
    1. Physical-based protocol
      1. RS232 Protocol
      2. RS485 Protocol
    2. Layered Based-protocols
      1. DNP 3.0
      2. ModBus
      3. IEC 61850
      4. LON
  4. Comparison of Relay Characteristics between Different Vendors

1. Comparison of Communication Medias

Let’s start with brief description of seven most known and most used communication medias used in power system communications (in terms of protection and automation):

1.1 Transmission Power Line Carrier

Economical, suitable for station to station communication. Equipment installed in utility owned area.Limited distance of coverage, low bandwidth, inherently few channels available, exposed to public access.

1.2 Microwave

Cost effective, reliable, suitable for establishing back bone communication infrastructure, high channel capacity, high data rates.Line of sight clearance required, high maintenance cost, specialized test equipment and need for skilled technicians, signal fading and multipath propagation.

1.3 Radio System

Mobile applications, suitable for communication with areas that are otherwise inaccessible.Noise, adjacent channel interference, changes in channel speed, overall speed, channel switching during data transfer, power limitations, and lack of security.

1.4 Satellite System

Wide area coverage, suitable to communicate with inaccessible areas, cost independent of distance, low error rates.Total dependency to remote locations, less control over transmission, continual leasing cost, subject to eavesdropping (tapping). End to end delays in order of 250 ms rule out most protective relay applications.

1.5 Spread Spectrum Radio

Affordable solution using unlicensed services.Yet to be examined to satisfy relaying requirement.

1.6 Leased Phone

Effective if solid link is required to site served by telephone service.Expensive in longer term, not good solution for multi channel application.

1.7 Fiber Optic

Cost effective, high bandwidth, high data rates, immune to electromagnetic interference. Already implemented in telecommunication, SCADA, video, data, voice transfer etc.Expensive test equipment, failures may be difficult to pin-point, can be subject to breakage.

Go back to contents ↑

2. Comparison of Different Communication Network Topologies

2.1 Point-to-Point network

Point-to-Point network is the simplest configuration with channel available only between two nodes.

Point-to-Point network - graphical model
Figure 1 – Point-to-Point network – graphical model
Suitable for systems that require high exchange rate of communication between two nodes.Communication can only be transferred between two nodes, disconnection of the communication channel will lead to a total loss of information exchange.

Go back to contents ↑

2.2 Star network

Star network consists of multiple point-to-point systems with one common data collector.

Star network - graphical model
Figure 2 – Star network – graphical model
Easy to add and remove nodes, simple in managing and monitoring, node breakdown does not affect rest of the system.The reliability of entire network depends only on single hub failure.

Go back to contents ↑

2.3 Bus network

Bus network has single communication path which runs throughout the system to connect nodes.

Bus network - graphical model
Figure 3 – Bus network – graphical model
Bus network is not dependent on a single machine (hub). This provides high flexibility in configuration (easy to remove or add nodes and node to node can be directly connected).High information load might delay the communication traffic speed. Also, it is sometime inefficient to utilize communication channels since the information cannot be exchanged directly between the desired relay and hub without passing through relays along the communication path. In other words, some relays may receive information packets which are unnecessary for them. Thus, it is also hard to troubleshoot the root cause of problem when needed.

Go back to contents ↑

2.4 Linear Drop and Insert network

Linear Drop and Insert network consists of multiple paths for relays to communicate with each other. Information between two non-adjacent nodes can be transferred directly passes through intervening node(s).

Linear Drop and Insert network - graphical model
Figure 4 – Linear Drop and Insert network – graphical model
When a certain communication channel drops, its bandwidth can be balanced by other channels.Lack of channel backup against fiber or equipment failure.

Go back to contents ↑

2.5 SONET Path Switched Ring

SONET Path Switched Ring comprises of two separate optical fiber links connecting all the nodes in counter rotating configuration.

In normal case, the information is transferred from A to C through outer ring (via B) which is the primary route (left figure). However if channel failure occurs, the information is transferred through inner ring which is secondary route (right figure).

SONET Path Switched Ring - graphical model
Figure 5 – SONET Path Switched Ring – graphical model
This type of network is redundant which means that channel failures will not affect the communication process.An unequal time delay between transmitter and receiver might cause the false operation of protective relays when there is a switch to from primary to secondary route in the case of channel failure.

Go back to contents ↑

2.6 SONET Line Switched Ring

SONET Line Switched Ring has the same structure as SONET Path type however one path is active and other is a reserved one.

Under normal condition, the active path transfers information via outer ring (left figure). However in case of channel failure, the inner ring is activated to reverse and transmit information through another direction (right figure).

SONET Line Switched Ring - graphical model
Figure 6 – SONET Line Switched Ring – graphical model
More efficient use of fiber communications for some applications.This communication type is not suitable for teleprotection applications since it requires complex handshaking (Synchronizing) that causes a delay of 60 ms.

Go back to contents ↑

3. Description of Different Communication Protocols

Communications protocols are sets of rules by which communication over a network is achieved. Communications protocols are responsible for enabling and controlling network communication.

Protocols set the rules for the representation of data, the signals used in communications, the detection of errors, and the authentication of computing devices on the network. It is not mandatory for relay manufacturers to follow the same protocols as shown in Anex table.

Communication protocols can be categorized into two groups which are (1) Physical-based protocols and (2) Layered-based protocols. Both types of protocol are briefly discussed in this technical article.

Go back to contents ↑

3.1 Physical-based protocol

Physical Based-protocols have been developed to ensure compatibility between units provided by different manufacturers, and to allow for a reasonable success in transferring data over specified distances and/or data rates.

The Electronics Industry Association (EIA) has produced protocols such as RS232, RS422, RS423 and RS485 that deal with data communications.

In addition, these physical-based protocols are also included in the “Physical layer” of the Open Systems Interconnection (OSI) model that will be explained in Layered-based protocols, section below.

Go back to contents ↑

3.1.1 RS232 Protocol

The RS232 Protocol is the most basic communication protocol which specifies the criteria for communication between two devices.

This type of communication can be simplex (one device acts as transmitter and other acts as receiver and there is only one way traffic i.e. from transmitter to receiver), half duplex (any of the device can act as a transmitter or receiver but not at the same time) or full duplex (any of the device can transmit or receive data at the same time).

A single twisted pair connection is required between the two devices. Figure 7 shows the RS232 protocol configuration.

RS232 Protocol configuration
Figure 7 – RS232 Protocol configuration

Go back to contents ↑

3.1.2 RS485 Protocol

This protocol is similar to the RS232 protocol which allows multiple relays (up to 32) to communicate at half-duplex. This half duplex scheme authorizes one relay either to transmit or receive command information.

This means that the information is handled by polling / responding.

The communication is always initiated by the “Master unit” (host) and the “Slave units” (relays) will neither transmit data without receiving a request from the “Master unit” nor communicate with each other.

There are two communication modes in RS485 protocol:

  1. Unicast mode and
  2. Broadcast mode

In the unicast mode, the “Master unit” sends polling commands, and only one “Slave unit” (assigned by an unique address) responds to its command accordingly. The “Master unit” will wait until it obtains a response from a “Slave unit”or abandon a response in case a pre-defined period expires.

RS485 Protocol configuration: Unicast mode
Figure 8 – RS485 Protocol configuration: Unicast mode

In the broadcast mode, the “Master unit” broadcasts message to all “Slave units”. Figure 8 and 9 show a simple RS485 protocol configuration in the unicast and the broadcast mode, respectively.

RS485 Protocol configuration: Broadcast mode
Figure 9 – RS485 Protocol configuration: Broadcast mode

Go back to contents ↑

3.2 Layered Based-protocols

Other protocols mentioned in Anex table are developed by the Open Systems Interconnection (OSI) model.

This model is a product of the Open Systems Interconnection (OSI) effort at the International Organization for Standardization.

How it works? – The model sub-divides a communication system into several layers. A layer is a collection of similar functions that provide services to the layer above it and receives services from the one below. On each layer, an instance provides services to the instances at the layer above and requests service from the layer below.

When data is transferred from one device to another, each layer would add the specific information to the “headers” and the information will be decrypted at the destination end.

Figure 10 demonstrates data communication using OSI model where “H” represents “headers”.

OSI model
Figure 10 – OSI model

Table 1 below describes function of each layer.

Table 1 – Functions of Open Systems Interconnection (OSI) model

Offers direct interaction of user with the software application. Adds an application header to the data which defines which type of application has been requested. This forms an application data unit. There are several standards for this layer e.g. HTTP, FTP, etc.
Handles format conversion to common representation data and compresses and decompresses the data received and sent over the network. It adds a presentation header to the application data unit having information about the format of data and the encryption used.
Establishes a dialogue and logical connection with the end user and provides functions like fault handling and crash recovery. It adds a session header to the presentation data unit and forms a session data unit.
Manages the packet to the destination and divides a larger amount of data into smaller packages. There are two transport protocols, Transmission Control Protocol (TCP) and User Datagram Protocol (UDP), in this layer.

Reliability and speed are the primary difference between these two protocols. TCP establishes connections between two hosts on the network through packages which are determined by the IP address and port number. TCP keeps track of the packages delivery order and check of those that must be resent.

Maintaining this information for each connection makes TCP a stateful protocol. On the other hand, UDP provides a low overhead transmission service, but with less error checking.

Controls the routing and addressing of the packages between the networks and conveys the packet through the shortest and fastest route in the network. Adds a network header to the Transport Data Unit which includes the Network Address.
Data Link
Specifies Physical Address (MAC Address) and provides functions like error detection, resending etc. This layer adds a Data Link Header to the Network Data Unit which includes the Physical Address. This makes a data link data unit.
PhysicalDetermines electrical, mechanical, functional and procedural properties of the physical medium.

Some of protocols, mentioned in Anex table, that are derived from OSI model are described below:

Go back to contents ↑

3.2.1 DNP 3.0

The Distributed Network Protocol (DNP) 3.0 is a protocol developed to achieve interoperability standard between substation computers.

This protocol adopts layers 1, 2 and 7 from the OSI model for basic implementation. A fourth layer (a pseudo-transport layer) can be added to allow for the message segmentation.

This DNP 3.0 protocol with a pseudo-transport layer is called the Enhanced Performance Architecture (EPA) model. It is primarily used for communications between master stations in Supervisory Control and Data Acquisition (SCADA) systems, Remote Terminal Units (RTUs), and Intelligent Electronic Devices (IEDs) for the electric utility industry.

This protocol does not wait for data as TCP/IP. If a packet is delayed, after a while, it will be dropped. This is because the protocol consists of embedded time synchronization (timetag) associated with messages.

This timetag’s accuracy is on the order of milliseconds. It is feasible to exchange messages asynchronously which is shown in a function of the polling/
response rate.

The typical processing throughput rate is 20 milliseconds.

Go back to contents ↑

3.2.2 ModBus

ModBus is also a three-layer protocol that communicates using a “master-slave” technique in which only one device (the master) can initiate transactions (called queries).

The other devices (slaves) respond by supplying the requested data to the master, or by taking the action requested in the query.

This protocol does not consist of embedded time synchronization as in case of DNP 3.0 that each message is stored in an internal buffer.

However, time synchronization can be implemented either using the external time synchronization source, such as Global Positioning System (GPS) or using the external timing mechanism, such as Inter-Range Instrumentation Group (IRIG) to keep Intelligent Electronic Devices (IEDs) in synchronism.

In general, IRIG provides accuracy in the 100 microsecond range but it requires dedicated coaxial cable to transport the timing signals which can be limitation for the number of connected devices (depending on cable length and device load).

On the other hands, GPS provides higher accuracy (in the range of 1 microsecond) compare to IRIG but cost and complications of antennas installation to every device are the restriction for the GPS deployment.

Nevertheless, the choice of time synchronization protocol is usually dictated by the number and type of power system devices as well as the physical arrangement of the equipment. The typical processing throughput rate of ModBus protocol is 8 milliseconds.

The protocol can be categorized into three frame formats which are American Standard Code for Information Interchange (ASCII), Remote Terminal Unit (RTU), and Transfer Control Protocol and Internet Protocol (TCP/IP) format.

The ModBus ASCII and ModBus RTU are both used in serial communication. The difference between these ASCII and RTU frames is the format of communication message. In the ASCII format, two ASCII characters are used in each 8 bit byte message whereas two 4 bit hexadecimal characters (or 8-bit binary) are used in case of the RTU format.

The advantage of ASCII format is that it allows time intervals of up to one second to occur between characters without causing an error.

On the other hand, the greater character density in the RTU allows better data throughput compare with the ASCII for the same baud (modulation) rate however each message must be transmitted in a continuous stream.

Figure 11 shows the Protocol Data Unit (PDU) for ASCII and RTU frame formats.

ModBus ASCII & RTU Protocol Data Unit (PDU)
Figure 11 – ModBus ASCII & RTU Protocol Data Unit (PDU)

Meanwhile, the ModBus TCP/IP is modified from the PDU frame with the Ethernet-TCP/IP as an additional data transmission technology for the ModBus Protocol.

First, an “Error Check” algorithm at the end of frame is removed and the Address Field (address of slave) is replaced by a new header called the ModBus APplication (MBAP) Header. This header consists of (i) Transaction Identifier, (ii) Protocol Identifier, (iii) Length Field, and (iv) Unit Identifier.

Figure 12 shows the Application Data Unit (ADU) for TCP/IP frame format (compare with PDU message).

The difference between ModBus and DNP 3.0 is the communication purpose. ModBus is suitable for communication within substations that are used for communicating with devices meant for protection control and metering.

Message frame comparison between ModBus PDU and ADU
Figure 12 – Message frame comparison between ModBus PDU and ADU

Meanwhile DNP 3.0 is suitable for communicate outside the substations (communication of data from substation to master control centers). This is because the ModBus protocol has limited function codes while the DNP 3.0 supports the specific data objects that provide more flexibility, reliability and security.

For example, the DNP 3.0 has ‘Control Function Code’ to perform specific function.

In addition, the ModBus protocol is a prototype for proprietary protocols such as K-BUS and SPA protocols which are of Areva-Alstom and ABB, respectively.

Go back to contents ↑

3.2.3 IEC 61850

IEC 61850 is an electrical substation standard promoted by the International Electrotechnical Commission (IEC).

The data models defined in IEC 61850 protocol can be mapped to various protocols, for example to Generic Object Oriented Substation Events (GOOSE) that allows for both analog and digital peer-to-peer data exchange (read more HERE.).

The protocol includes time tags and also messages that can be exchanged asynchronously.

The typical processing throughput rate is 12 milliseconds.

IEC 61850 provides many advantage over other protocols such as programming can be done independent of wiring, higher performance with more data exchange, or data is transmitted multiple times to avoid missing information.

Go back to contents ↑

3.2.4 LON

The Local Operating Network (LON) protocol equates all seven layers of the OSI Model. It is capable of establishing network communications not only for power system applications, but also for factory automation, process control, building networks, vehicle networks etc.

This may be considered as a drawback in relay communication perspective since the LON protocol occupies seven layers in order to transfer information, thus it provides lower data exchange rates compare to the EPA model such as DNP 3.0.

Go back to contents ↑


Comparison of Relay Characteristics between Different Vendors

Vendors (their communication methods)
RS232; RS485;
IEC 61850; ModBus
TCP/IP; DNP 3.0;
IEC 60870-5-104
RS232; RS485; IEC
IEC 60870-5-103;
LON; SPA; DNP 3.0;
SEL; ModBus
RS232; RS485;
ModBus; IEC
DNP 3.0;
IEC 61850
RS232; RS485;
DNP 3.0; ModBus
Voltage protection
RS232; RS485;
IEC 61850;
ModBus TCP/IP;
IEC 60870-5-103
RS232; RS485; IEC
IEC 60870-5-103;
LON; SPA; DNP 3.0;
EIA 485; ModBus
RTU; EIA 232

Go back to contents ↑

Reference // Power System Protective Relaying: basic concepts, industrial-grade devices, and communication mechanisms by Rujiroj Leelaruji and Dr. Luigi Vanfretti (KTH Royal Institute of Technology – Electric Power Systems Department)

SEARCH: Articles, software & guides

Premium Membership

Premium membership gives you an access to specialized technical articles and extra premium content (electrical guides and software).
Get Premium Now ⚡

About Author


Edvard Csanyi

Electrical engineer, programmer and founder of EEP. Highly specialized for design of LV/MV switchgears and LV high power busbar trunking (<6300A) in power substations, commercial buildings and industry fascilities. Professional in AutoCAD programming. Present on


  1. Eddy Gudijanto
    Jun 07, 2018

    The paper above is very useful, I would be grateful if you could share with us.
    Many thanks

  2. Marc André
    Jun 06, 2018

    Please send me this content’s PDF version in my e-mail address

  3. Durga Prasad
    Jun 05, 2018

    Please send me this content’s PDF version in my e-mail address

  4. Rammohan Reddy C
    Jun 05, 2018

    This is really super to learning for the networking.You could send networking related topics. I really likes….

  5. Srijan Ganguly
    Jun 04, 2018

    Please send me this content’s PDF version in my e-mail address

Leave a Comment

Tell us what you're thinking... we care about your opinion!

Premium Membership

Get access to premium electrical guides, technical articles and much more!
(20% Discount Code: PREMIUMEEP6723)