Optimise and reap
Richard Schimmel describes how ‘off-site’ acceptance testing can be optimised from a utility and system integrator point of view using features of the IEC 61850 design standard.
In the current energy market, utilities operate the grid increasingly closer to its limits. For this to be possible, more substations need to be automated or retrofitted in a short period of time, and the scheduled outage time need to be reduced. To do more work in the same time period with the same, or even less, personnel, the utility needs to reduce testing efforts and time on site. Accomplishing this is feasible when much of the acceptance test is executed in the factory or test laboratory. With IEC 61850, the testing in the factory can be optimised and even automated by using an advanced automated acceptance test system.
The IEC 61850 standard, Communications Networks and Systems in Substations for substation automation, is a worldwide accepted single, standard solution for communications integration. The standard:
Uses self-description and object modelling technology (logical nodes) to simplify the integration and configuration process for the user
Incorporates robust, very high-speed control communications messaging (GOOSE) that can operate among relays and other IEDs to eliminate panel wiring and controls
Includes a vendor-independent substation configuration language (SCL) to specify all communications and data models of each intelligent electronic device (IED) of the substation automation system.
These features are used to reduce the testing effort by automating the acceptance test in the factory or test lab.
Quality assurance
The factory acceptance test (FAT) is part of the total quality assurance process. The IEC 61850-4 standard identifies the following quality assurance stages (See figure 1).
The manufacturers develop and type test their devices according to ISO 9001. As a final proof of quality, the IEC 61850 interface of each type of device must be certified according to IEC 61850-10 conformance test procedures. Presently, over 80 devices from various manufacturers have been certified – these certificates can be downloaded from the UCA international users’ group website. In March 2008, the UCA international users group approved conformance test procedures for HMI and gateway systems. Certificates are expected soon.
Project execution and acceptance testing
The system integrator can visit the website and select certified devices and systems. They can then ‘integrate’ multiple devices and systems from one or more vendors into substation automation and protection system based on the specification from the utility. By selecting certified devices and systems, the system integrator has assurance that the communications between these devices are all according to the standard. The system integrator will use a substation configuration tool to develop the "Substation Configuration Description (SCD)" file in SCL format. The SCD file includes the full communications configuration and IED data models of the SAS. The system integrator will then perform a system integration test to verify if the SAS meets the requirements from the utility. This verification is also part of the quality assurance process, but is less visible to the end-user. When the integration test of the complete SAS is successful, the system integrator can invite the utility to start the formal "off-site" FAT.
During the FAT, the system integrator must prove to the utility that the SAS satisfies its requirements. In this stage, an advanced test system is required to make this task both efficient and effective. It is a well-known fact that it is much easier to resolve detected test issues in the factory than it is on-site. With an advanced test system, much more of the SAS functionality can be tested "off-site," and as such, unpleasant surprises on-site will be prevented. In case an issue is detected during the FAT, the system integrator can repair it and the system integrator and utility can perform a re-test. This process is repeated until all test issues are resolved and all test cases are passed. The test results should convince the utility to accept and install the SAS on-site. When the SAS is completely tested "off-site," all that must be done onsite is the hardwiring to the switchyard and the connection to the real SCADA system in the control centre. As a result, the on-site testing – the site acceptance test (SAT) – can be reduced to a "point-to-point" test from the hardwiring to the control centre. This causes onsite time, risk and testing efforts to be significantly reduced.
Acceptance test system
As mentioned, the test system should be both effective and efficient to support the acceptance testing. Now we will investigate what this means in terms of test system requirements. To be both efficient and effective, the test system should be automated. In case test issues are detected, the applicable test cases or in the worst case, the complete test needs to be repeated. An automated test system enables:
Reproducible test results independent of by whom, when or where the test is executed
Fast test execution - the test system is faster than typing/clicking of a human operator
Automatic analysis of test results to indicate whether a test passed or failed
Automatic archiving of test results
No human intervention during a test – no way to interfere or disturb the test results
Test engineers with limited knowledge of IEC 61850 or other communication protocols can execute the test.
The scope of the test is the complete SAS. In case the SAS is not complete, the missing connections and/or IEDs need to be simulated as later described. As such we can consider the SAS a black box. The only links to/from the black box SAS are the wiring to the switchyard and the control centre connection. The control centre connection may be IEC 60870-5-104, DNP3.0 or, in the near future, IEC 61850.
As a minimum, the test system should simulate both the switchyard and the control centre connection. The testing procedures shall be described in terms of inputs/outputs to/from the switchyard, and events to/from the control centre. For performance test cases, the timestamp accuracy of the test system is about one ms.
Other test system requirements are:
Scalable in number of inputs/outputs
Simple and fast wiring from the I/O system to the SAS
Simple and fast configuration of the I/O system
Simple and fast editing of the test scripts
Simple and fast configuration of the test system control centre connection
Archive and display test results in a utility acceptable open format like Comtrade.
CIGRE workgroup B5.32 has developed a draft report regarding the requirements of a functional test system. The above-mentioned requirements are a subset of these workgroup specifications. Figure 3 provides a schematic overview of a test system based on hardwired inputs/outputs to the switchyard.
The test system consists of the following parts:
Test master and test scripts; archive and display test results, configure test scripts, and
Process simulator; simulate the switchyard, e.g. switch delay time and volts/amps on open/closing switches
Network simulator; simulate the power flow to/from the substation
Test scheduler; start/stop one or more test scripts
Test timer; start/stop time-related operations and measure time differences
Test arbiter; decide if test case passed or failed based on configured performance criteria and logic
Operator simulator; simulate operator control such as opening/closing a switch.
I/O driver; driver(s) for I/O system(s)
HMI simulator; to simulate the HMI when not yet available
CC simulator; simulates the control centre connection
Analyser; monitor and analyse IEC 61850 messages on the substation LAN
IED simulator; simulate one or more missing IEDs
I/O system; conversion to/from digital data to binary/analogue electrical signals
Test scripts
The only challenge for such test systems will be simple script editing and substation configuration. For IEC 61850, all communications are specified within the SAS. This excludes the connection to the control centre. In case the control centre connection is 104, the SCD file may be extended with the 104 parameters as specified in the new IEC 61850-80-1 report. The process simulation component can use the single line scheme from the "substation system description" component in the SCD file.
To synchronise the time, the test system and I/O system use the same time source (GPS clock). The test system will include a set of common test scripts. To prevent script editing for each substation, the events are configured as parameters outside of the script. Consequently, the same script can be used many times with different parameters. Most parameters may be automatically extracted from the SCD; the test engineer enters the remaining parameters. When necessary, the user of the system adjusts or creates new test scripts for the non-common test cases.
The typical sequence of a test script is:
Test connection; connect to I/O system and gateway
Test set-up; initialise all values of the I/O system and gateway for a test
Test start, initiate a sequence of operator control and/or switchyard events while running the process simulation and network simulation
Test stop; collect results including events from the gateway link, events from I/O system and GOOSE virtual events on the substation LAN
Test disconnection; disconnect from I/O system and gateway Test verdict; compare if the time difference of events are within the limit and/or the logic result is true.
Using the test system
There may be limitations regarding the availability of equipment in the factory or test lab. It may also be necessary to start the testing when not all IEDs are present and/or it is not practical to have many copies of the same (feeder) bays. In these cases, the SAS is not complete. To enable the testing, the missing IEDs need to be simulated by using an additional IED simulation test tool and the SCD file. In the worst case scenario, when only the substation HMI or gateway is present, it can already be tested by simulating all IEDs of the entire substation. When IEDs become available they simply replace the simulated IED. Such substation, or multi-IED simulator, has been developed by KEMA and is already in use by several vendors. When only one or more IED’s are present and the HMI or gateway is missing, the test system needs an additional ‘HMI simulator.’
The acceptance test system cannot be fully automated. In some test cases, manual interaction of the test engineer is required, e.g. disconnecting a LAN cable or entering a control command at the substation HMI. In these cases the test system can ask the test engineer to perform a certain action or enter a result. For example "Disconnect the link between Ethernet switch 3 and 4: Ok" or "Verify circuit breaker X position on the HMI is intermediate: Yes/No".
Whenever a test case fails, it will be necessary to know what/where it went wrong inside the SAS. Especially important are the GOOSE messages between the IED’s and the messages between the IED’s and the gateway or HMI system. To monitor and analyse these messages on the substation LAN, a protocol analyser is required. Considering the large number of messages communicated by a remote controllable analyser with build-in automatic error detection, display of SCL semantics, advanced GOOSE filtering and Comtrade export features will significantly speed-up the uncovering of the issue source. The UniCA 61850 analyser satisfies these requirements. Recently, the Dutch utility "Eneco Infra" performed a market research study on IEC 61850 analysers and concluded that the UniCA analyser is the only analyser to satisfy all its technical requirements.
KEMA has gathered extensive knowledge and experience by conformance testing many IEC 61850 and IEC 60870-5-104 devices/systems from all major manufacturers. The automated test system KEMA uses for conformance tests is very similar to the acceptance test system described above. As a result, KEMA expects to make the acceptance test system available relatively soon.
The described automated acceptance test system can perform a wide range of test cases. By using test scripts, the test can be automated and is very flexible - from automated point-to-point testing to performance testing of multiple distributed functions under stress. By using such a test system during the off-site acceptance testing, the system integrator and utility significantly reduce both test efforts and scheduled outage time. Utility or system integrators will benefit the most in cases where they expect to execute multiple SAS retrofit projects in a short period of time.
The author works at the KEMA Protocol Competence & Test Centre. In 1995 he developed a test system for IEC 60870-5-101 and 103. He is an active member of IEC TC57 WG10 and the UCA international users group, product manager of the UniCa family of test tools, and responsible for IEC 61850 consulting, testing and certification services. He may be contacted through Rolf.vanStenus@kema.com. Netherlands-based KEMA (www.kema.com) is an independent company that is active worldwide in the energy chain, specialising in high-level services in the field of business & technical consultancy, operational support, measurements & inspections, and testing & certification. In the ME region, KEMA is responsible for a large number of high-level projects, such as executing a feasibility study and realising of a conceptual power supply scheme design for ‘The World’ in Dubai, and optimising maintenance and related operating processes of the electricity transmission and distribution network in Bahrain. KEMA’s ME office is based in Dubai.
References
IEC 61850, Communications Networks and Systems in Substations, First edition, 10 parts in 14 standards documents issued between 2001 and 2005, International Electrotechnical Commission, Geneva, Switzerland, http://www.iec.ch
UCA international user group – Test register of IEC 61850 certified devices: www.ucaiug.org/TESTarchive/UCAIug%20Test%20Docs
UCA international user group - Conformance Test Procedures for Client System with IEC 61850-8-1 interface, March 2008
Functional Testing of IEC 61850 Based Systems, CIGRE Work Group B5.32, expected December 2009
Market research on IEC 61850 Protocol analysers (in Dutch), June 2008, Eneco Infra, Rotterdam, the Netherlands
KEMA Protocol Competence & Test Centre, www.kema.com/pctc
About Automation Insight
Automation Insight is a complimentary monthly publication designed specifically for the utility industry and those serving the utility industry. For comments or suggestions on future article topics, please e-mail automation.insight@kema.com.
Subscribe to KEMA's Automation Insight newsletter.
Automation Insight is an opt-in subscription. KEMA does not sell or otherwise make public subscriber information and honors all ‘unsubscribe’ requests. To unsubscribe, please e-mail automation.insight@kema.com with 'unsubscribe' in the subject line.