ALSTOM T&D. Network Protection And Automation Guide (NPAG)

Подождите немного. Документ загружается.

Alstom Grid 24-1

Chapter 24

The Digital Substation

24.1 Introduction

24.2 Communications

24.3 Cyber Security

24.4 Current and Voltage – Digital Transformation

24.5 IEC61850

24.6 Conclusion

24.7 References

24.1 INTRODUCTION

The Digital Substation is a term applied to electrical

substations where operation is managed between distributed

intelligent electronic devices (IEDs) interconnected by

communications networks. It became possible by using

computing technology in the substation environment. In the

third edition of the Protective Relays Application Guide, first

printed in June 1987, a new chapter was introduced: “The

application of microprocessors to substation control”.

Microprocessors were introduced into substation products

such as protection devices to improve the performance of the

main product functions. Features such as improved accuracy

and stability were delivered. Communications ports were also

incorporated, but these were more like a developer’s debug

tool than the sophisticated ones we might recognise today.

Communication facilities developed, however. They were

refined to provide connection to SCADA equipment to reduce

hardwiring. Communications could provide operators with

informative interfaces using software packages running on

personal computers. Significantly they developed to enable the

transmission of data between the different substation devices.

As computing power became greater and cheaper, it became

possible to integrate multiple functionalities into single devices,

resulting in fewer devices being required to implement the

same traditional functionality.

As a result of functional integration and communications, the

traditional divisions between protection, monitoring and

control became blurred. The term ‘substation automation’

was becoming used by protection engineers. This is reflected

in a later revision to PRAG: In July 2002 PRAG was replaced

by the first edition of NPAG and “The application of

microprocessors to substation control” was removed in favour

of new chapters associated with “Automation”.

Automation offered exciting possibilities, and improvement in

communications and computing resulted in more sophisticated

IEDs for substation automation becoming available.

Along a similar time line, innovative approaches for

instrumentation transformers were being developed, with so-

called merging units converting indicative signals of power

system quantities into data streams for communication to

IEDs.

The technology was now available to communicate current

copyright notice but may not be copied or displayed for commercial purposes without the prior written permission of Alstom Grid.

Network Protection & Automation Guide

24-2

and voltage information to the different IEDs within the

substation on a so called 'process bus', whilst the IEDs

communicated information between themselves and the

control centre on a so called 'station bus'. The digital

substation was close to becoming a reality, but despite

attempts at standardisation, proprietary solutions hampered

the end goal of "plug-and-play".

The breakthrough came with the introduction and adoption of

IEC61850, an open standard for electrical substation

automation. This standard caters for data modelling, station

buses, process buses, etc., facilitating seamless

interoperability.

The digital substation brings major benefits in terms of design

and engineering, installation, and operation. Off-the-shelf

solutions can be offered, modifications can be easily

accommodated, cabling (and hence costs), are reduced, and

embedded diagnostics assure system integrity.

This chapter takes a tour through the enablers of the digital

substation, looking at communications (and how to ensure

that they are sufficiently secure, reliable, and dependable),

current and voltage transformation coupled with merging

units, and concluding with an introduction to the standard

IEC61850.

24.2 COMMUNICATIONS

Communication is the conveyance of information from a

source to a destination. In 1949 C E Shannon and W Weaver

[24.1] produced a mathematical model for communications as

shown in Figure 24.1.

Figure 24.1: The Shannon Weaver model

The model shows that for the information to transfer from the

source to the destination, a transmitter (encoder) and a

receiver (decoder), both connected by a channel are required.

It also demonstrates that sources of noise can interfere with

the process. Considering communication between two people,

the source is the brain of the person who wants to get the

message across, the transmitter could be that person’s voice

and the channel is the atmosphere. At the receiving end, the

receiver would be the other person’s ears and the destination

that person’s brain. Audible noise could drown out the

message causing it not to be received and this is a possible

cause of communication breakdown, but it is not the only one.

If the source is speaking but the destination is not listening (i.e.

they are not synchronised), communication won’t be

successful. Similarly, if the two people talking don’t share a

common language, and/or if certain protocols are not observed

(e.g. not both speaking at the same time), failure is inevitable.

Successful communication requires common understanding

between source and destination.

24.2.1 The OSI Model for Computer Communications

In the context of substation automation, communication is the

transfer of information from one computing device to another

but, as with the example above, problems with

synchronisation, language, and protocol can all cause

communication failure.

The substation communication might be in the form of a

dedicated link between two devices, or it may be over some

form of communications network.

The International Standards Organisation (ISO) recognised the

need for a framework for inter-device data communications,

and in 1984 introduced the Open Systems Interconnection

(OSI) model.

The OSI model divides the data communication process into

seven distinct layers. Each of the seven layers defines how the

data is handled during the different stages of transmission.

Each layer provides a service for the layer immediately above

it.

A model representing the seven layers is shown in Figure 24.2

with the purpose of the different layers being described below.

Link layer

Presentation layer

Application layer

Session layer

Transport layer

Network layer

Physical layer

Communication application

Layer FunctionData Unit

Data

Segment

Packet

Frame

Bit

Media

Layers

Host

Layers

Layer

Type

Encryption and data representation

Inter-host communication

End-to-end connections and reliability

Logical addressing

Physical addressing

Media, signal and binary transmission

Figure 24.2: The OSI seven layer model

24.2.1.1 OSI Layer 1 – The Physical Layer

Every data message is transmitted on some medium. This

medium usually takes the form of cables, wires, or optical

fibres, but it could just as well be wireless, with the data being

carried on electromagnetic waves. The physical layer specifies

the type of medium to be used between one end of the data

exchange and the other, the type of connectors to be used and

the voltage and current levels, and if applicable, optical

copyright notice but may not be copied or displayed for commercial purposes without the prior written permission of Alstom Grid.

Chapter 24  The Digital Substation

24-3

characteristics defining the state of the data bits. A very

common medium, particularly in Ethernet systems started

with the CAT 5 UTP cable (Category 5 Unshielded Twisted

Pair), which consists of four colour-coded twisted pairs of

wires terminated with RJ45 connectors. As the name implies,

this type of cable is not shielded against electromagnetic

interference. There is also a version of this cable called CAT 5

STP (Category 5 Shielded Twisted Pair) that has a metallic

outer sheath, providing electromagnetic shielding. This type of

cable is often used in noisy environments like the substation.

CAT 5, has more recently evolved into new standards such as

CAT 6 and CAT 7, which offer even more robust shielding, fire

protection etc.

24.2.1.2 OSI Layer 2 – The Data Link Layer

The physical layer provides the Data Link layer with bits. The

Data Link layer now provides some intelligence to this

sequence of bits by defining Data Frames. These Data Frames

are packets of data containing the data to be transmitted and

some control information governing the transmission. The

control information comprises flags to indicate the start and

end of the message. The standards used at this layer must

ensure that the control flags are not mistaken for data and that

the data frames are checked for errors. An example is given in

Figure 24.3.

Figure 24.3: Example of a data frame

A commonly used Data Link protocol is Ethernet. Examples of

equipment that work at this layer are Ethernet switches and

bridges.

24.2.1.3 OSI Layer 3 – The Network Layer

The network layer is concerned with packet delivery. Logical

paths are established between the sending and receiving

equipment by adding information onto the data frame which

defines where the packet has come from, and where the

packet is going to. This takes the form of a logical source and

destination address for each packet of information. A

commonly used Network protocol is IP (Internet Protocol).

An example of equipment that operates at the network layer is

a router.

24.2.1.4 OSI Layer 4 – The Transport Layer

The transport layer is the first layer that is not concerned with

the mechanics of the data transfer. It is concerned with

managing and sequencing the packets once they have arrived.

There is an array of transport layer protocols ranging from the

very simple, which simply accepts the data as it comes in, not

caring about whether the packets have errors nor even if they

are in the right order, through to quite complex protocols

which check the data for errors, send out an acknowledgement

to the sending equipment, order the packets into the correct

sequence and present the data to the next layer guaranteed

error-free and correctly sequenced.

An example of a simple protocol is UDP (User Datagram

Protocol). A more sophisticated widely used transport layer

protocol used is TCP (Transmission Control Protocol).

Layer 4 functionality is achieved by the devices at either end of

the transmission path, which is usually a computer. In the

context of substation protection, this would be the IED.

24.2.1.5 OSI Layer 5 – The Session Layer

The first four layers establish a means of reliable

communication between two IEDs (computers), but they do

not deal with intelligent management of the communication.

This is performed at the session layer. It is the first layer that

has user interaction. Each communication session is governed

by criteria pertinent to the session. One communication

session may be downloading a file from a web site, whilst

another may be working on a file situated on a remote server.

Session layer software can implement password control,

monitor system usage, and allow a user interaction with the

communication.

24.2.1.6 OSI Layer 6 – The Presentation Layer

As the name implies, the presentation layer concerns itself with

how the data is presented. A good example of this is ASCII

(American Standard Code for Information Interchange). An

ASCII code is an 8-bit binary code, which defines the familiar

character set used to produce text. For example the character

‘A’ is defined by the binary code 01000001, which is 41

hexadecimal or 65 decimal. The presentation layer, if

compliant with the ASCII format, can interpret this.

24.2.1.7 OSI Layer 7 – The Application Layer

The application layer is the layer that interacts with the user.

This is in the main software that allows the user to define the

communication. Examples of application level protocols would

be file transfer protocol (FTP), hypertext transfer protocol

(HTTP), post-office protocol 3 (POP3), or simple mail transfer

protocol (SMTP).

copyright notice but may not be copied or displayed for commercial purposes without the prior written permission of Alstom Grid.

Network Protection & Automation Guide

24-4

24.2.2 Communications Between IEDs

The OSI model outlines the need for compatibility between all

communication layers between computers (which are

generally in the form of IEDs in the substation). This section

explores the lowest two layers; the physical and data link

layers.

24.2.2.1 Physical Connection to IEDs for Substation

Control and Automation

In substation control and automation systems, connection to

IED communications ports at the physical layer (OSI layer 1) is

generally to one of three standards:

x EIA 232

x EIA 485

x Ethernet

EIA (Electronic Industries Association) is a set of standards

defined for connecting electronic devices together. They were

formerly

known as the recommended standards (RS) and

often people will still refer to them as RS 232 and RS 485.

Note: Although, traditionally the standards were designed

with metallic connections between devices, more recent times

have seen an increase in optical fibre as the medium

connecting networks and devices. Many of the physical layer

and data link layer protocols have now been modified to work

over optical fibre as well.

EIA 232

EIA 232 is an electrical connection allowing full duplex

communication between two devices. Prior to the

development of the USB port, it was the typical serial port

found on most computers and is characterised by Table 24.1

Item Details

Max. number of transmitters 1

Max. number of receivers 1

Connection type 25 core shielded

Mode of operation DC coupling

Maximum distance of transmission 15m

Maximum data rate 20kbits/s

Transmitter voltage 5V min, 15V max

Receiver intensity 3V

Driver slew rate

30V/

Table 24.1: EIA(RS)232 characteristics

EIA 232 was designed to allow computers to connect using

Modems and is suitable only for point-to-point connection. It

uses switched single-sided 12V (nominal) signals for data

transmission as well as handshake signals to control the

communication. Due to Modem limitations, it is not generally

used at speeds in excess of 9,600 bits per second. No isolation

is specified and so it is only suitable for connection over very

short distances. For permanent connection in a substation

environment, the use of optical isolating units should be

considered to avoid damage caused by induced transients. If

transmission over longer distances is required, some form of

EIA 232 to fibre-optic converter and a fibre-optic

communications link should be used. Since EIA 232 provides

only point to point connectivity it is not used in automation

systems, rather it is limited to relaying information to a SCADA

system.

EIA 485

EIA 485 is an electrical connection characterised by Table

24.2.

Item Details

Max. number of transmitters 32

Max. number of receivers 32

Connection type Shielded Twisted Pair

Mode of operation Differential

Maximum distance of transmission 1200m

Maximum data rate 10Mbits/s

Transmitter voltage 1.5V min

Receiver intensity 300mV

Table 24.2: EIA(RS)485 characteristics

EIA 485 provides a two-wire half-duplex connection designed

for multi-drop connections as shown in Figure 24.4 which

makes it more suitable than EIA 232 for use in automation

schemes.

Figure 24.4: Multi-drop connection of RS485 devices

EIA 485 uses differential signalling on twisted pairs and can be

isolated. The multi-drop connection (sometimes called daisy-

chaining) generally has a limit of 1km, and although in theory

the number of devices connected (nodes) is not limited, a

practical limit is normally 32 per bus, with repeaters being

used if further expansion is required. Communication speeds

of 64kbits per second (kbps) can be reliably achieved. EIA 485

allows relatively simple networks to be constructed very

efficiently and cost effectively.

Ethernet

Ethernet is a standard, which defines the connection of

computing devices to local area networks (LANs). As per

copyright notice but may not be copied or displayed for commercial purposes without the prior written permission of Alstom Grid.

Chapter 24  The Digital Substation

24-5

IEEE 802.3, standard signalling speeds are 10Mbps, 100Mbps

or 1Gbps. It is the most widespread LAN technology. The

specification allows connection to be made either electrically

using an RJ45 connector, or by direct fibre-optic connection.

Ethernet over fibre-optic cables provide a mechanism for

extremely high speed and noise resistant communication,

making it the ideal communication medium for the substation.

Ethernet is discussed in detail, later in the chapter.

24.2.2.2 Network Topologies

When linking multiple products together to form a computer

network, in addition to the physical connection of the device, it

is also necessary to consider the network topology as well as

the protocol or language by which information is exchanged.

For automation systems to be effective, information must be

communicated reliably between different devices in the

system. In early centralised systems, a hierarchical series of

connections is required for devices at the acquisition and

process levels to communicate data upwards and to receive

commands in return. This tree-type structure typifies the kind

of connection that the point-to-point EIA 232 communications

can bring.

The multi-drop capabilities of EIA 485 do more to encourage

distribution of control since the EIA 485 network will have

devices connected on the same multi-drop bus, with a master

connected to many slaves.

Systems based on Ethernet technology provide more scope for

different topologies by virtue of equipment such as switches,

hubs, bridges and routers.

24.2.3 Serial Communications

This is a form of communication whereby bits of data are

exchanged serially through the signalling channel. With

parallel communications multiple bits of data are exchanged in

parallel. An example of parallel communication is the LPT

printer port that used to feature on computers, printers, and

office equipment before the widespread deployment of

Ethernet in the working environment.

EIA 232, EIA 485, and Ethernet are all forms of serial

communication. The much higher speed of Ethernet sets it

apart from the others. As a consequence, in the domain of

substation automation, communications based on EIA 232

and EIA 485 tend to be referred to as “serial” communications,

with Ethernet being singled out for separate attention. This

chapter adheres to that convention and continues with a

discussion on “serial” communications based on EIA 232 and

EIA 485, later followed by Ethernet communication.

24.2.3.1 Serial Communications Protocols

A communications protocol is a set of standard rules for data

representation, signalling, authentication and error detection,

which defines the transfer of information over a

communication channel. Put another way, for devices to be

able to speak to each other, they need to share a common

language and rules of engagement.

When digital communication facilities were first added to

computer based protection devices, no standard protocol

existed for this communication; manufacturers developed their

own proprietary solutions to exploit the benefits of the

communication interfaces. An example of a proprietary

protocol is the Courier protocol that was developed by the

former GEC Measurements (now Alstom).

A substation control system is required to communicate with

all the distributed functionality in the system. If devices from

different vendors are to be included in the system, then the

different protocols will need to be supported. This increases

the system engineering work and consequently the cost.

Consider a relatively simple system, with a bay controller

connected to IEDs from different manufacturers A, B, and C as

depicted in Figure 24.5.

Figure 24.5: Multi-protocol system

If the three products all contain key elements to implement the

control system, then either the bay controller would need to

support three communications protocols (as shown in the

figure), or some form of protocol converter devices would need

to be included. Either way, it can be seen that the use of

different protocols increases the engineering time, and the

component costs associated with implementing the system.

From the systems' engineering perspective, as well as the

utilities' perspective, the adoption of a standard protocol by the

manufacturers brings clear benefits.

Efforts were made to develop a standard solution and from the

protocols available, three open standards for IED serial

communications emerged:

copyright notice but may not be copied or displayed for commercial purposes without the prior written permission of Alstom Grid.

Network Protection & Automation Guide

24-6

x MODBUS

x IEC 60870-5-103

x DNP3

Modbus

Modbus was published in 1979 by Modicon. Originally

designed for programmable logic controllers of the

time, it is a master-slave protocol that allows a master

to read or write bits in registers of the slave devices. It

can be implemented on simple serial connections (EIA

485), but has also been migrated to Ethernet.

IEC 60870-5-103

IEC 60870-5-103 was developed by Technical

Committee 57 of the International Electro-technical

Committee (IEC TC57). It is based on, and is a

superset of, the German VDEW communications

standard. Primarily serving European clients, it defines

the standards for communication between protection

equipment and the devices of the control system. As

well as simple electrical connection, the standard also

defines a direct communication interface implemented

over optical fibre.

DNP3

DNP3 was developed by Harris in the 1990s. It drew

on early work of TC57 on the IEC 60870-5 (before it

was standardised) to develop something with a specific

focus on the North American market that could be

quickly implemented. Like Modbus, it is also supported

on Ethernet as well as simple serial connection.

Although originally targeted at the North American

market, it has been deployed elsewhere.

The use of these open communications standards simplified

the system engineering, making future upgrading and

expansion easier, hence reducing overall life-cycle costs.

In general, the open standards of Modbus, IEC60870-5-103

and DNP3 cover layers 1-4 of the OSI model. They allow

automation systems engineers to mix and match IEDs from

different manufacturers with the same protocol, satisfying the

clients’ needs for protection and control better than if a mix of

proprietary protocols were used. Each of the protocols has

particular advantages, but they all suffer disadvantages too:

x Although they are standard protocols, all have

ambiguities that result in different implementations in

different IEDs.

x The IED data model (i.e. the

way the data representing

the IED is structured and presented over the

communications link) varies between vendors, and can

sometimes var

y between different versions of the same

IED from the same vendor.

x Different functionalities are supported by the different

protocols, and different IED vendors may support

different levels of those functionalities.

x The provision of private codes in the IEC 60870-5-103

protoc

ol permits much greater functionality, but at the

same time hinders interoperability of equipment from

different vendors because there is no need for the

format of such messages to be made public. In effect,

the use of ‘private’ messages by vendors of devices

essentially turns the standard into several proprietary

ones.

Due to continuing developments in the fields of computing and

communications, there is a need for a truly open

communications architecture allowing ‘plug-and-play’

connectivity between devices from different manufacturers for

effective, efficient substation automation. The IEC61850,

standard, based on Ethernet satisfies that need.

24.2.4 Ethernet Communications

Ethernet defines the connection of computing devices to local

area networks (LANs). It is the most widespread LAN

technology. Standardised in IEEE 802.3, it describes the

requirements for network connection according to layer 1 (the

physical layer) and layer 2 (the data link layer) of the OSI

model.

At the physical layer the specification allows connection to be

made either electrically using an RJ45 connector, or by direct

fibre-optic connection using LC or ST (otherwise known as

BFOC2.5) type connectors.

Standard signalling speeds for Ethernet communications are

10Mbps, 100Mbps, 1Gbps and 10Gbps.

At the data link layer (layer 2 of the OSI model), MAC (Media

Access Control) addresses take care of the physical addressing

of devices on a network. These physical addresses are set in

the manufacturing process. MAC uses a 48 bit addressing

scheme allowing a unique address for every piece of

equipment manufactured.

The most commonly used network protocol (layer 3 of the OSI

model) is the Internet Protocol (IP), and the most commonly

used transport layer mechanism (OSI layer 4) is the

Transmission Control Protocol (TCP). Often these are put

together as 'TCP/IP' and are synonymous with Ethernet and

the internet.

At the network layer, IP is used to transfer packets of data on

the network and across network boundaries from the source to

copyright notice but may not be copied or displayed for commercial purposes without the prior written permission of Alstom Grid.

Chapter 24  The Digital Substation

24-7

the destination. To do this, devices are assigned IP addresses

and these addresses have two purposes: The first is to identify

the host network (which may be subdivided into sub-

networks), whilst the second is to identify the device on the

network or sub-network (subnet). The IP address is a 32 bit

binary number, which can be written in the form

nnn.nnn.nnn.nnn, where nnn represents a binary octet, and is

an integer value between 0 and 255. A typical IP address is,

therefore, 192.168.0.2. Each device on the network will have

an IP address that is unique and may be configured by a

network server, a system administrator, or by an address auto-

configuration. The most significant bits of the address are

used for network identification and are known as the routing

prefix. The least significant bits are known as the rest field and

provide unique identification of the device on the local network

to which it is connected. Typically the two most significant

octets identify the network. The two least significant octets are

used for device identification, but can also be used to create

subnets. A so called subnet mask is then used to identify the

subnet address. For example if the third octet were used as a

subnet identifier, then performing a logical AND between the

IP address and 255.255.255.0 (the subnet mask) would

identify that, in the case above, the subnet address is

192.168.0.0 and the device is at address 2 on that subnet.

When a packet of data is placed on the IP network, it is sent to

its destination network by means of routers, switches and

hubs. These devices are discussed in the next section.

24.2.4.1 Ethernet Hubs, Switches and Routers

An Ethernet hub is, as the name suggests, a hub for linking

multiple Ethernet devices together. Hubs work at the physical

layer. Devices on the network transmit data packages to the

hub and the hub passes them to all connected devices. There

is little arbitration and data collision is common. As such, they

are not appropriate for use in mission critical applications such

as substation automation.

Ethernet switches operate at the data link layer. Ethernet

switches are intelligent devices that provide a common

connection point for devices on a network. They store

incoming packages and forward them to the appropriate

destination on the LAN when the channel is free. Full duplex

operation provides for consistent, reliable operation. Figure

24.6 shows a typical Ethernet switch outline.

Figure 24.6: Ethernet Switch outline (4-port)

Ethernet routers are used as gateways between the LAN and

wide-area networks (WANs). They are similar to switches, but

they have fewer ports and act at the network layer. The

information is presented to the router in the form of a

sequence of IP packets. Each packet has a logical source

address and destination address associated with it, and the

router passes the packet on to the next step of the destination.

Packets with the same destination address do not necessarily

follow the same route. If one route gets congested, for

example, the router may decide to send the packet onwards

via an alternative route (rather like a motorist may choose to

avoid road works). For this reason, packets making up a

message may arrive at their destination out of sequence. In

many cases, it is critical that the message arrives error-free

and in the correct sequence. Higher level protocols in the

computer, such as TCP, ensure that this happens.

24.2.4.2 Ethernet Topologies

The topology of a network is the way in which devices are

connected together. Figure 24.7 shows the two simplest

topologies which are the star and the ring so-called because

they resemble stars and rings when drawn.

Figure 24.7: Simple star and ring network topologies

In the star topology, a physical connection runs from each

device on the network to a central location, usually an Ethernet

switch.

In the ring topology, a physical connection is daisy-chained

around the devices in the form of a ring.

Ethernet is inherently a star-based network topology. Initially

copyright notice but may not be copied or displayed for commercial purposes without the prior written permission of Alstom Grid.

Network Protection & Automation Guide

24-8

it did not allow ring topology connection due to the risk of

traffic getting stuck in an endless loop around a ring and the

consequent paralysis of the network as the amount of stuck

traffic increased. IEEE 802.1D introduced the Spanning Tree

Protocol (STP) which was developed to overcome this

limitation, and allow Ethernet to be used in rings as shown in

Figure 24.8.

Figure 24.8: Ethernet topology with STP

STP was limited in terms of speed, so IEEE 802.1W Rapid

Spanning Tree Protocol (RSTP) was developed from STP to

provide faster performance and hence more effective Ethernet

ring topologies.

24.2.4.3 Principles of Redundancy in Communications

networks

The term “redundancy” can be open to misinterpretation as it

wrongly implies that something may not be needed. In the

context of communication networks, what it actually means is:

“Redundancy is any resource that would not be needed if there

were no failures”. Redundancy is therefore a provision of

transparent backup. It is required where failure cannot be

tolerated, such as critical applications like transmission

substation automation.

Figure 24.9 shows how redundancy can be incorporated into a

star network. A connection from each node goes to a different

switch, providing an alternative path. This topology is called

Dual Homing Star Topology.

Figure 24.9: Redundant connections in star topology

Figure 24.10 shows simple redundancy with a ring topology.

The cable is daisy-chained from device to device in a ring, the

idea being that if the link from one direction fails, the link in

the other direction can be used to effect transactions. In the

event that there is a break at one point of the ring, appropriate

redundancy protocols can automatically readjust the ring such

that the data is sent back in the opposite direction, ensuring it

will get to its destination.

Device Device

Device

Root switch

Break in

ring

Data sent

back in

other

direction

Data sent

back in

other

direction

Figure 24.10: Redundancy in a ring topology

This inherent redundancy is one of the features that RSTP

offers.

In this example, an alternative path is provided in a 'standby'

arrangement. When the primary path fails, this is detected and

the alternative path is switched into action. This type of

redundancy is known as dynamic redundancy, and it needs a

certain amount of time to operate.

Where an alternative path is provided in a 'workby'

arrangement such as that shown in Figure 24.9, it is called

static redundancy. Should one path fail, the alternative paths

are continually working and, should one path fail, no switching

time is lost because the other path is already operating.

copyright notice but may not be copied or displayed for commercial purposes without the prior written permission of Alstom Grid.

Chapter 24  The Digital Substation

24-9

24.2.4.4 Redundancy requirements for substation

automation networks

Dynamic redundancy is cost-effective, but requires switchover

time. Static redundancy provides seamless switchover but

generally costs more.

The Dual Homing Star, and Rapid Spanning Tree protocols

both offer redundancy, but neither presents an ideal solution

for the needs of substation automation. To overcome the

limitations, the standard IEC 62439-3 (Industrial

communication networks - High availability automation

networks) was developed for substation automation.

Of particular interest is IEC 62439-3 clause 4 which describes

a Parallel Redundancy Protocol (PRP), and IEC 62439-3

clause 5 which describes High-availability Seamless

Redundancy (HSR).

Switchover delays and grace time are critical to the substation

automation system. The switchover delay is dictated by the

grace time and is the most constraining factor in fault-tolerant

systems. The grace time is the time that the plant allows for

recovery before taking emergency actions.

Some examples of grace times for automation systems are

given in IEC62439-3 and are shown in Table 24.3.

Applications Typical grace time

Uncritical automation applications, e.g. enterprise systems 20s

Automation management, e.g. manufacturing 2s

General automation, e.g. process automation, power plants 0.2s

Time-critical automation, e.g. synchronised drives 0.02s

Table 24.3: Examples of grace time

Table 24.4 provides examples of recovery times for various

protocols used for implementing redundancy.

Protocol Description

Frame

Loss

Typical

recovery time

IP IP routing Yes 30 seconds

STP Spanning Tree Protocol Yes 20 seconds

RSTP Rapid Spanning Tree Protocol Yes 2 seconds

CRP

Cross-network Redundancy

Protocol

Yes 1 second

MRP Media Redundancy Protocol Yes 200 milliseconds

BRP Beacon Redundancy Protocol Yes 8 milliseconds

PRP Parallel Redundancy Protocol No 0 seconds

HSR High-availability Seamless Ring No 0 seconds

Table 24.4: Typical recovery times for common redundancy protocols

From Table 24.3 and Table 24.4 it can be seen that systems

based on PRP and HSR are appropriately suited to the needs

of substation automation networks, providing true static

redundancy. This is also known as 'bumpless' redundancy.

24.2.4.5 IEC 62439-3 PRP

PRP is standardised in IEC62439-3 (clause 4) for use in Dual

Homing Star Topology networks. PRP is capable of providing

bumpless redundancy for real-time systems, and hence

becomes the reference standard for star-topology networks in

the substation environment.

A PRP compatible device has two ports operating in parallel,

each port being connected to a separate LAN segment.

IEC62439-3 assigns the term DANP (Doubly Attached Node

running PRP) to such devices.

Figure 24.11 shows an example of a PRP network. The doubly

attached nodes DANP 1 and DANP 2 have full node

redundancy, whilst the singly attached nodes SAN 1 and SAN

4 do not have any redundancy. Singly attached nodes can,

however, be connected to both LANs, via a device that

converts a singly attached node into a doubly attached node.

These devices are referred to as redundancy boxes or

RedBoxes. Devices with single network cards such as PCs,

printers, etc., are singly attached nodes that may be connected

into the PRP network via a RedBox as shown in figure Figure

24.11.

LAN A Switch LAN B Switch

DANP 2RedBoxDANP 1

SAN 1 SAN 4SAN 2 SAN 3

Figure 24.11: Example PRP redundant network

Both ports share the same MAC address, so there is no impact

on the functioning of the address resolution protocol. Every

data frame is seen by both ports.

When a node sends a frame of data, the frame is duplicated on

both ports and thus on both LAN segments, providing a

redundant path for the data frame in the event of failure of one

of the segments. When both LAN segments are operational, as

is the normal case, each port receives identical frames and

these identical frames need to be carefully handled. The

handling brings overheads, but bumpless redundancy is

assured.

24.2.4.6 IEC 62439-3 HSR

HSR is standardised in IEC62439-3 (clause 5) for use in Ring

copyright notice but may not be copied or displayed for commercial purposes without the prior written permission of Alstom Grid.

Network Protection & Automation Guide

24-10

Topology networks. HSR is capable of providing bumpless

redundancy for real-time systems and becomes the reference

standard for ring-topology networks in the substation

environment.

HSR works on the premise that each device connected in the

ring is a doubly attached node running HSR. IEC62439

assigns the term DANH (Doubly Attached Node running HSR)

to such devices. As per PRP, singly attached nodes are

connected via a so-called Redbox.

Figure 24.12 shows a simple HSR network, where a doubly

attached node is sending a multicast frame (that is a frame

that is intended for multiple recipients on the network). The

frame (C frame) is duplicated, and each duplicate frame is

tagged with the destination MAC address and the sequence

number. The frames differ only in their sequence number,

which is used to identify one copy from another. For

convenience, the duplicate frames are labelled the A frame and

B frame. Each frame is sent to the network via a separate port.

The destination DANH receives two identical frames from each

port, removes the HSR tag of the first frame received and

passes this to its upper layers. This now becomes the D frame.

The duplicate frame is discarded.

The nodes forward frames from one port to another unless the

particular node is the node that originally injected it into the

ring.

DANH

Source

DANH Redbox

DANH DANH DANH

A frame B frame

C frame

Switch

Singly Attached

Nodes

D frameD frame

D frameD frameD frame

Figure 24.12: HSR for multicast traffic

With unicast frames (frames that are intended for a single

destination), there is just one destination and the frames are

sent to that destination alone. All non-recipient devices simply

pass the frames on. They do not process them in any way. In

other words, D frames are produced only for the receiving

DANH. This is illustrated in Figure 24.13

Figure 24.13: HSR for unicast traffic

The use of PRP and HSR delivers network reliability, but the

networks also need to be secure. Security within network

communications comes under the banner 'Cyber Security',

which will be discussed in the next section.

24.3 CYBER SECURITY

When communications were first introduced into the

substation environment for control and automation, they were

fully contained within the substation and based on proprietary

protocols. As a consequence they enjoyed inherent security.

They were “secure by isolation” (if the substation network is

not connected to the outside world, it can’t be accessed from

the outside world), and “secure by obscurity” (if the formats

and protocols are proprietary, it can be very difficult, to

interpret and hack into them).

The increasing sophistication of substation protection, control

and automation schemes coupled with the advancement of

technology and the desire for vendor interoperability has

resulted in standardisation of networks and data interchange

within substations. In wide-area applications, substations can

be interconnected with open networks such as corporate

networks or the internet, which use open protocols for

communication. Open protocols mean that the security that

obscurity brought cannot be assumed, and interconnection via

open networks means that the security that isolation brought

them cannot be assumed either. This leaves the networks

vulnerable to so-called cyber-attacks.

Cyber-security provides protection against unauthorised

disclosure, transfer, modification, or destruction of information

and/or information systems, whether accidental or intentional.

Factors to be considered in the context of cyber-security

include:

 Confidentiality (preventing unauthorised access to

information).

copyright notice but may not be copied or displayed for commercial purposes without the prior written permission of Alstom Grid.

ALSTOM T&amp;D. Network Protection And Automation Guide (NPAG)

Подождите немного. Документ загружается.

ALSTOM T&D. Network Protection And Automation Guide (NPAG)