Practical aspects of IEC 61850-9-2 implementation in microprocessor-based protection and control IEDs

Protection and control intellectual electronic devices (IEDs) respond to the signals of currents and voltages measured at certain points of the power system, and assess the state of the protected power system component.

Conventionally, analog values are injected directly into the IED through instrument transformers (Fig. 1). IEDs combine analog-to-digital conversion of the signals with their analysis (digital filtering) and decision-making algorithms.

The IEC 61850 standard, which becomes more and more popular, allows digital exchange of data between (standalone) merging units and protection devices. Merging units (MUs) (Fig. 2), being an integral part of the digital substation, repeatedly digitize the analog signals and transfer them to the process bus as packets ensuring labeling and integrity of data during the transfer [1], [2]. The process bus represents a communication network, interconnecting data publishers and subscribers. Protection and control IEDs receive digital packets and process currents and voltages. The transfer of the instantaneous values of currents and voltages is required for the performance of all range of functions available in the relays. Thus, protection and control devices connected to the process bus register electric processes as sampled signals of currents and voltages in the same way as if the analog-to-digital conversion was carried out directly in the relay.

Figure 1. Conventional approach of connecting IEDs to instrument transformers
Figure 1. Conventional approach of connecting IEDs to instrument transformers

The digital communication is generally carried out through optic cables and offers a number of advantages for the secondary circuitry of the substation as compared to the conventional wiring [3], [4], [5]:

 

  • considerable reduction of cabling at the substation,
  • advanced electromagnetic interference immunity of data transmission,
  • ensuring galvanic isolation of circuits,
  • advanced reliability of data transmission (control over the availability of the communication channel and data integrity),
  • improved measurement accuracy,
  • a possibility of incorporating up-to-date measurement
    methods. At the same time, there are a few reasons for limited use of MUs, such as:
  • a single sampling frequency (dictated by the most demanding devices),
  • MU measurement errors influence all of its subscribers,
  • great demand for the computing capacity of the protec- tion and control IEDs for input data processing,
  • difficulties in time synchronization of measurements.
Figure 2. Providing digital data to protection and control IEDS over process bus network
Figure 2. Providing digital data to protection and control IEDS over process bus network

The transition to digital technologies retains and partially steps up the requirements for the reliability and compatibility. This paper focuses on a study of the operation modes of communication systems based on the transmission of sampled values in comparison with the conventional technologies, while also exposing flaws in the solutions and ways to diminish them.

Classification of digital communication effects for protection and control IEDs

Conventionally, microprocessor-based relays take samples of analog signals at certain pre-defined points in time. The sampling rate may be constant, though frequency tracking can also be implemented. Since IEC 61850-9-2 requires that all measurements be taken by the DMT, the relay has no control over the stream of input data. According to IEC 61850-9-2LE, the sampling rate of 4000 Hz is to be utilized for a 50 Hz power system [6]. Such peculiarities may provoke a number of issues typical for the systems with digital signal transfer and suspending normal operation:

  • loss of one or several packets of digital data in digital communication channels;
  • delay of packet delivery through the process bus caused by some operation peculiarities of the communication equipment;
  • loss of connection with the MU as a result of the communication channel breakdown or MU failure;
  • poor data quality due to MU failure, detected by its self-test system;
  • time synchronization errors.

At the same time digital communication is free of issues typical for the conventional transmission of analog signals, such as:

  • noise induction by other analog chan- nels caused by the lengthy cabling;
  • impulse noise induction during light- ning as well as during switching opera- tions;
  • damage of the current and voltage secondary circuits between the measuring transducer and relays;
  • damage of relay circuits, processing input analog signals.

IEC 61850-9-2 packet loss scenario

The distinctive feature of digital communication is encoding information as discontinuous values. This is why there is a possibility of damaging part of a packet or even losing it (for example, due to the overload of communication channels). According to IEC 61850-9-2LE, the packet transmitted includes one sample of each of the three phase currents and three phase voltages, as well as current and neutral voltage. Most filtering algorithms are designed for equal distribution of samples on the time axis and are very sensitive to the loss of even one of them. Under such conditions the relay must take special actions like disabling certain functions or computing the sample missed. As the former may have an adverse impact on the protection performance, the latter is seen to be more appropriate. With that in mind, let’s go through the following methods of information recovery:

  1. Setting the missed sample to zero.
  2. Use of the previous reading.
  3. Linear interpolation of the adjacent data points.

There are also more accurate methods of interpolation which require advanced calculations and use a larger data window. They are not included in this report.

Figure 3. Methods of lost data recovery: curve 1 – original signal; curve 2 – zeroing out the lost reading; curve 3 – use of the previous reading; curve 4 – restoration of the lost sample through linear interpolation of the adjacent data points.
Figure 3. Methods of lost data recovery: curve 1 – original signal; curve 2 – zeroing out the lost reading; curve 3 – use of the previous reading; curve 4 – restoration of the lost sample through linear interpolation of the adjacent data points.

Each of the considered methods of data recovery is depicted in Fig. 3, where curve 1 represents the original signal, curve 2 – zeroing the missed sample out, curve 3 – use of the previous reading, and curve 4 – linear interpolation of two adjacent points to restore the lost one.

Figure 4. illustrates the accuracy of the linear interpolation for a power system transient. The original signal (curve 1) is plotted with the y-axis on the left (axis 1), and the error of the lost sample recovery calculated for each of the successive data points (curve 2) is plotted with the y-axis on the right (axis 2).
Figure 4. illustrates the accuracy of the linear interpolation for a power system transient. The original signal (curve 1) is plotted with the y-axis on the left (axis 1), and the error of the lost sample recovery calculated for each of the successive data points (curve 2) is plotted with the y-axis on the right (axis 2).

The error of each recovery method has been estimated assuming the loss of one sample per cycle of a 50 Hz sine wave. Table 1 presents the maximum deviations of the computed phasors (found through Fourier analysis) from the true one.

Table 1. Estimation of errors associated with each of the recovery methods
Table 1. Estimation of errors associated with each of the recovery methods

The advantage of the algorithm based on the linear interpolation is clearly seen.

The peculiarities of relay operation mentioned above should be taken into account when designing protection algorithms as well as when determining the relay settings. The loss of samples has the greatest impact on the operation of differential protection and on protection functions, based on incremental quantities.

Irregular packet delivery from merging unit

Another issue of the digital data transfer is a packet delivery time drift when data are transferred across the network. This leads to irregularities in the flow of data to the protection equipment, which is uncommon for protection algorithms requiring, as a rule, strict execution order. There are two ways to eliminate this problem: irregular execution of the protection algorithms inside IED or their delayed execution. In the first case, there may be new issues resulting from the need to ensure accurate time triggering. In the second case, to cope with packet delays in the communication channel and to keep the required regularity of data processing, a buffer of samples is created, thus enabling delayed data delivery to the protection functions. Buffer time should provide for the maximum possible time of data delivery in the communication channel. The overall relay tripping time then increases by this delay.

Data reception from several merging units

When a relay receives streams from several merging units, their time synchronization is required. To time-align the samples received from different sources, a sample counter is used, which is transferred in the data packet. The counter is incremented by one with each measurement and sets back to zero every second.

In the buffer of sampled values, where they are kept until supplied to the relay algorithms, the received data can be aligned (Fig. 5). The buffer accounts for some delay in the arrival of packets as long as it is within a preset boundary.

Figure 5. Aligning samples from different MUs
Figure 5. Aligning samples from different MUs

If the measuring equipment is not synchronized, the collected samples cannot be correctly aligned. The joint use of such data is impossible. The relay functions are then blocked, and an alarm condition is indicated to the maintenance personnel.

It is important to ensure synchronous time-stamping of the samples. This is achieved by the use of special time synchronization system, providing the accuracy of ±1-4 µs.

If the moments of taking the samples do not coincide, the computational error increases when using the data from different sources.

Loss of connection with merging unit

The loss of connection with a data source for the process bus in terms of its impact on the protection relay performance is similar to damage of current and voltage secondary circuits. Such a loss is detected after the packet delay exceeds a preset threshold, which enables a timely reaction of the relay. The relay functions requiring data from the failed source are then blocked, and the loss of connection is indicated to the maintenance personnel.

In order to minimize the chances of such a scenario, the communication channels redundancy is implemented (Fig. 6).

Figure 6. Communication channels redundancy
Figure 6. Communication channels redundancy

The need for process bus channels redundancy is caused by the requirement of their high reliability. The process bus is actually the main way of interaction between a relay and the external switching and measuring equipment. The application of such redundancy methods as STP (Spanning Tree Protocol), RSTP (Rapid Spanning Tree Protocol), and MRP (Media Redundancy Protocol) does not comply with the requirements imposed by the bus. For example, in the RSTP protocol, the connection restoration time may amount to one second and more. Even in case of a ring-type network topology, the restoration time below 100 ms may hardly be reached. The MRP protocol allows reducing this time down to 10-50 ms, which is still too much for the process bus.

The redundancy problem may be solved by the application of the advanced methods of parallel and seamless redundancy – PRP (Parallel Redundancy Protocol) and HSR (High-availability Seamless Redundancy) [7]. The distinctive feature of these methods is that at least two simultaneous active connections between the devices exist at all times. All the frames are synchronously sent and received by the devices through two Ethernet channels. In case of two identical parallel networks, switching between the primary and secondary channels can be made in almost no time.

It is important to have built-in support for this protocol in protection and control IEDs and MUs themselves. Although specialized RedBox devices can be utilized if the relays and MUs do not support the protocol, such a solution would make the overall network structure more complicated, introduce additional delays into the data transfer path, and, which is also important, eliminate the opportunity to provide backup for the Redundancy Box itself, which would not favor the reliability of data transmission.

Poor data quality

Merging units shall perform the diagnostics of their state. In case of internal failure, samples are marked with a poor quality attribute according to IEC 61850, and each measurement has its own quality attribute.

If a poor quality sample is received, the original value can be restored by means of the algorithms applied to a lost sample calculation. The restoration algorithms are only utilized for those signals in a packet that are marked as faulty, while the other received signals are not modified. If more than one poor quality sample is received, protection and control functions depending on this data are blocked, and the relay informs the maintenance personnel about the failure.

Use of external MUs allows implementing the data source redundancy and improving the system operation reliability at a low cost. In such a case, instead of the corrupted data, the relay in some cases may use the samples received from a good quality redundant source, which facilitates the uninterrupted protection of the power system component without having to take the relay out of service. Such switching can be performed automatically, without a delay.

Conclusion

The digital transmission of analog signals has its strengths and weaknesses. The required reliability of protection and control applications is achieved through the extension of diagnostic possibilities as well as through simplification of providing redundancy for both communication channels and data sources.

In order to overcome the undesired effects accompanying the digital transmission of analog signals, one has to implement special algorithms in IEDs, along with taking into account additional requirements when designing the process bus and time synchronization system at a substation as well as when configuring the protection relays.

References

  1. IEC 61850-1 (International Standard). Communica- tion Networks and Systems in Substations – Part 1: Introduction and overview.
  2. IEC 61850-7-2 (International Standard). Communication Networks and Systems in Substations – Part 7-2: Basic communication structure for substation and feeder equipment – Abstract communication service interface (ACSI).
  3. Gorelik T.G. (NIIPT OJSC), Kirienko O.V. (Energopro- mAvtomatizatsia Ltd.) Power facilities automation using the “Digital Substation” technology. The first Russian pilot model// Relay protection and automation. – 2012. – No. 3. – p. 86- 89.
  4. I.N. Dorofeev (LISIS Ltd.). New principles of designing automation systems on substations with “digital” secondary circuits. Organization of control and protection systems based on the iSAS software package// RELAVEKSPO-2012. Collection of abstracts and reports. – p. 116-121.
  5. А. Boyko, Y. Yershov, Y. Mikhaylenko. DIGITAL CURRENT AND VOLTAGE CIRCUITS. Application in relay protection. // News in Electrical Technology. – 2006. – No. 6(42).
  6. IEC 61850-9-2 LE (Lite Edition). Implementation Guideline for Digital Interface to Instrument Transformers using IEC 61850-9-2.
  7. IEC 62439-3 (International Standard). Industrial Communication Networks – High Availability Automation Networks – Part 3: Parallel Redundancy Protocol (PRP) and High-availability Seamless Redundancy (HSR).

Digital Substation

(close)