Tiza Group

Home / Guides / Protection, control and communications

IEC 61850-5 Message Types and Performance Classes P1-P3

The seven message types and the P1 to P3 and M1 to M3 performance classes of IEC 61850-5 edition 1: 10 and 3 ms, integrity and system load.

By Shadai García, CEO and protection engineer at Tiza Group.

senderreceiverIEDIEDSender stacktaGridtbReceiver stacktc3 ms (P2/P3)10 ms (P1)Ed. 1 · transfer time = ta + tb + tcMessage typesspeed1A · 1Bfast messages2medium speed3low speed4raw data5file transfer6time synchronization7commands with access controlpowercontrolcommunicationTT0 to TT6: edition 2TransferSenderGridReceiverTime© tizagrp.comsenderreceiverIEDIEDSender stacktaGridtbReceiver stacktc3 ms (P2/P3)10 ms (P1)Ed. 1 · transfer time = ta + tb + tcMessage typesspeed1A · 1Bfast messages2medium speed3low speed4raw data5file transfer6time synchronization7commands with access controlpowercontrolcommunicationTT0 to TT6: edition 2TransferSenderGridReceiverTime© tizagrp.com
IEC 61850-5 message types and performance classes: transfer time and types 1A to 7. Location in the system and operating principle. Illustrative: real values come from the nameplate and the manufacturer's manual.
Short answer

Edition 1 of IEC 61850-5 groups the data exchanged into seven message types, from 1 (fast and trip) to 7 (commands with access control), and defines performance classes for control and protection (P1, P2 and P3) and for metering and power quality (M1, M2 and M3). For tripping, class P1 sets 10 ms and P2 and P3 set 3 ms of total transfer time. Classes TT0 to TT6 belong to edition 2 of the same part, not to edition 1.

Why classify the messages

Thousands of individual data items are exchanged between the logical nodes of a substation, the PICOMs (see PICOM in IEC 61850-5). Many are alike: all trips, for example, have almost identical requirements apart from their source. That is why part 5 condenses them into seven message types and subdivides some into performance classes. The type describes the communication requirements of a message; because it is defined per message and not per installation, it is independent of the size of the substation.

This article describes edition 1 of part 5 (2003). That edition does not use classes TT0 to TT6; those belong to edition 2 and are explained in transfer time in IEC 61850.

What is measured: the transfer time

The standard defines transfer time as the complete transmission of a message, with the necessary handling at both ends. It is counted from the moment the sender puts the content at the top of its communication stack until the receiver takes it out of its own. It therefore includes three segments: the communication processor of the sender, the network (with waits and passage through routers and other devices) and the processor of the receiver.

The standard adds a practical warning: any test and verification of the transfer time must be done at site acceptance, because the end devices and the network devices may come from different manufacturers.

The seven message types

TypeNameTypical contentTotal transmission time
1AFast, tripTrips; also interlocks, transfer trips and discrimination logic between protective devices10 ms (P1); 3 ms (P2 and P3)
1BFast, otherClose, reclose command, start, stop, block, unblock, record trigger, change of state100 ms or less (P1); 20 ms (P2 and P3)
2Medium speedNormal states and a single measured value, with the sender’s time stampless than 100 ms
3Low speedComplex messages: slow automation, event records, reading and changing settings, time-stamped alarms and events, non-electrical measurements such as temperatureless than 500 ms
4Raw dataOutput of digital transducers and instrument transformers: continuous, synchronized streamsdefined by the trip time
5File transferLarge record, settings and information filesno specific limit; typically 1000 ms or more
6Time synchronizationSynchronization of the internal clocks of the IEDsthe resulting accuracy is required
7Commands with access controlCommands from the local or remote operator with greater security; based on type 3 with a password or verificationas for type 3

Type 6 is subdivided into 6a for the station bus and 6b for the process bus, in the table of part 1. Type 7 messages are mandatory for all communication using interface 7, the interface for external technical services, and file transfer requests from a remote access must be of type 7. When passing down to the process level, those commands can become messages with type 1 properties.

Part 5 also associates each type with the interfaces where it usually appears (numbering in levels and interfaces): type 1B on IF3, IF5 and IF8; type 2 on IF3, IF8 and IF9; type 4 on IF4 and sometimes IF8; and type 7 on IF1, IF6 and IF7.

Performance classes: P1, P2 and P3

There are two independent groups of classes, one for control and protection and one for metering and power quality. Within one substation not all links have to meet the same class: the station level and the process level can be chosen separately, and within the process level each bay can use a different class, depending on the quantity and rated value of the equipment.

ClassTypical application
P1Distribution bay, or bays where low requirements are accepted
P2Transmission bay, or when the customer does not specify otherwise
P3Transmission bay with the highest-performance synchronization function and breaker differential

For tripping (type 1A), the standard reasons in power-system cycles: class P1 asks for something on the order of half a cycle, and therefore defines 10 ms; classes P2 and P3 ask for less than a quarter of a cycle, and therefore define 3 ms. For the other fast messages (1B), P1 allows up to 100 ms, and P2 and P3 on the order of one cycle, set at 20 ms.

Classes M1, M2 and M3 are for metering and power quality:

  • M1: revenue metering with accuracy class 0.5 (IEC 62053-22) and 0.2 (IEC 60044-8), up to the 5th harmonic.
  • M2: revenue metering with class 0.2 and 0.1, up to the 13th harmonic.
  • M3: power quality measurement up to the 40th harmonic.

Raw data and synchronization

For type 4, the standard tabulates the resolution and the sampling rate per class. In protection and control: P1 with 480 samples per second and 13 bits, P2 with 960 and 16 bits and P3 with 1920 samples per second and 18 bits for current (16 bits for voltage). In metering: M1 with 1500 samples per second, M2 with 4000 and M3 with 12,000. These rates are functional requirements of edition 1; the way the samples are actually sent is covered in sampled values.

For type 6 the standard defines time synchronization accuracy classes:

ClassAccuracyUse
T1±1 msEvent time stamping
T2±0.1 msZero crossings, distributed synchrocheck, synchronized switching
T3±25 µsRelated to P1
T4±4 µsRelated to P2 and M1
T5±1 µsRelated to P3 and M2 and M3

The standard clarifies that these accuracies are functional requirements and that the implementation may require synchronizing the clocks with an order of magnitude more precision. It also states that it does not define direct requirements for the synchronization messages, except for the resulting time accuracy in the system. On substation clocks, see time synchronization. What IEC 61850-10 measures in GOOSE, SV, report and command latency is explained in IEC 61850-10 Performance Testing.

How a performance class is reached
  1. PICOM. Each data item has a source, a destination and a timing requirement
  2. Message type. PICOMs with similar requirements are grouped into one of the seven types
  3. Performance class. Depending on the bay and function, P1, P2 or P3 is chosen (or M1 to M3 for metering)
  4. Transfer time. The requirement is verified end to end, stacks and network included

Data integrity

Integrity means that, with a given background noise, the resulting errors stay below an acceptable limit. Part 5 refers to the three integrity classes of IEC 60870-4: all safety messages, such as commands and trips with a direct effect on the process, must have the highest class, 3; the others can use a lower one, but not below 2. The known measures to achieve it are three: suitable design of the equipment and the network (for example, protective enclosures and optical fiber), appropriate coding, and the use of sequences of at least two steps, such as select before operate for commands. The standard leaves their application outside part 5, but requires them to be taken into account when modeling the services and defining the mappings.

System requirements

Part 5 also explains how to check that the times are met. There are two methods: the PICOM method, which calculates the data flow from a base of functions and information and considers only the content, without headers, and the local network simulation method, which does include message overhead, addressing, multicast and collisions and delivers average and maximum delays. The first serves to determine total rates; for total transfer times, simulation or an equivalent must be used.

Some conclusions from the examples annex of edition 1: in the cases studied, abnormal or emergency loads on the station bus add between 50 % and 100 % to the normal load, and at the process level there is no notable difference because the devices send constant streams. Very different substations can give similar loads, and others that are almost the same can give very different loads, because the allocation of logical nodes and the network layout weigh heavily. Grouping data objects into messages reduces traffic noticeably. The standard’s examples use the technology of the time, so their absolute figures are not useful for sizing a current network: the useful part is the method. For that, see Sampled Values and GOOSE bandwidth.

What to do with this in a project

  • Assign each signal in the list a message type and a class; it is the basis for sizing the network and for writing the test requirements.
  • Distinguish the transfer time (stack to stack) from the total clearing time, which also includes the relay, the contact and the circuit breaker.
  • If a document cites TT classes, confirm which edition of part 5 it refers to.
  • Plan the verification of the transfer time for the acceptance tests, with the final equipment and network; see standards for FAT and SAT.

IEC 61850 timing tests with Tiza Group

Do you need to check that trips and blocking by message meet the time your project requires? At Tiza Group we verify it in the FAT and SAT of your substation or plant, at medium, high and extra-high voltage, in Colombia or elsewhere on request. In the quotation we recommend only the tests and equipment you really need.

References

  • IEC 61850-5 (edition 1, 2003): clauses 13 to 15 and annexes B, C and I.
  • IEC TR 61850-1 (edition 1, 2003): clause 6.2, table 1.

Frequently asked questions

Where do the 3 ms and 10 ms for tripping by message come from?

From edition 1 of IEC 61850-5, for the trip message (type 1A). Class P1 sets about half a cycle, that is 10 ms. Classes P2 and P3 set less than a quarter of a cycle, that is 3 ms. It is the total transfer time between the communication stacks of the two devices, not the operating time of the circuit breaker.

What is the relationship between classes P1, P2 and P3 and classes TT0 to TT6?

They are two ways of classifying the same requirement in different editions of part 5. Edition 1 uses seven message types with P and M classes; edition 2 expresses the transfer time in classes TT0 to TT6. In both, tripping is the most demanding requirement: 3 ms for P2 and P3 in edition 1 and class TT6 in edition 2.

What integrity does the standard require for a trip or a command?

The highest, class 3 of the three integrity classes taken from the IEC 60870-4 series. All safety-related messages, such as commands and trips with a direct effect on the process, must use that class. The others can use a lower one, but never below class 2.

WhatsApp