Testing of NCIT Based Protection Systems

Authors: Michael Freiburg, OMICRON electronics and Alexander Apostolov, USA

The testing of the merging units (MU) or remote units is the next step in the process. It is required to ensure that all components of the bus protection system are functioning as specified. If a directional comparison bus protection system is used, it is important to properly simulate the transition between pre-fault to fault condition. While the testing of conventional bus protection systems requires the simulation of currents only, the testing of the directional functions in the remote units requires simulation of both currents and voltages.  Each of the above described applications imposes different requirements for the tools used for testing of the protection system. At the same time the requirements for testing are affected by the type of test being performed - type testing, acceptance testing, commissioning or troubleshooting.

IEC 61850 has significant impact on the way distributed bus protection systems can be designed and implemented. 

Non-Conventional Instrument Transformer Interfaces (NCIT)

Non-conventional instruments can be of different types and with different interfaces. At the beginning, they were typically available with low voltage analog interface, but with the advancement of communications technology the trend shifted towards digital interface that was first specified by IEC TC 38 in IEC 60044-8 and now in the IEC 61869 series the low power instrument transformers are specified in -6 to -15. With the development of IEC 61850 9-2 and the implementation agreement known as IEC 61850 9-2 LE, it became clear that substation protection systems based on digital communications offer significant advantages over conventional hard wired solutions.

Conventional Protection Interface: The conventional application shown in Figure 1 connects the secondary of the current and voltage instrument transformers with copper cables to the analog inputs of the protection relays located in the substation control house. Conventional hard-wired protection systems have several deficiencies:

  • Susceptible to CT saturation, mainly due to the lead resistance
  • High costs of copper cables
  • Requirements for maintenance testing
  • Limited flexibility
  • Requirements for outages in cases of changes in the zone of protection
  • Safety issues related to open CT circuit conditions

The fact that conventional instrument transformers have been creating safety concerns (explosions, open CT circuits) and require separate protection and metering instrument transformers has been pushing utilities to take another look at using the different types of non-conventional instrument transformers available today. Next to the above deficiencies, conventional instrument transformers are still a preferred option for many utilities due to the vast experience with these units regarding operation and testing, the long lifetime, robustness and stability. However, a better understanding of NCITs and their proper commissioning, testing and maintenance will help to fully benefit from these technologies as usage will be likely increasing.

Low-power NCIT interface: This technology uses a low-power analog interface which has restricted usability for protection applications, and with the advancement of communications technology the concept of the Merging Unit (MU) - a device with an analog input and digital output - was introduced. This changed the interface of the protection and other devices in the substation (Figure2). This interface provides all the benefits that are typical for sampled values based interfaces with multifunctional protection IEDs. Like all other such interfaces it is preferred that the MU is located as close as possible to the low-power NCIT output to reduce the effect of substation transients on the operation of the protection functions.

Process bus based protection applications offer important advantages over conventional hard wired analog circuits that are important in the case of bus protection. The first very important one is the significant reduction in the cost of the system due to the fact that multiple copper cables are replaced with a small number of fiber optic cables.

Using a process bus also results in the significant reduction in the possibility for CT saturation (if still conventional CTs are being used) because of the elimination of the current leads resistance.

Process bus based solutions also improve the safety of the substation by eliminating one of the main safety related problems - an open current circuit condition. It becomes non-existent if optical current sensors are used.

Last, but not least, the process bus improves the flexibility of the bus differential protection, because any changes will only require modifications in the subscription of the protection IEDs receiving the sampled analog values over IEC 61850 9-2.

NCIT with direct IEC 61850 9-2 LE interface:  Depending on their operating principle (for example optical CTs), some NCITs with embedded MU may have direct IEC 61850 sampled values interface as shown in Figure3. Such an interface simplifies significantly the design of the system, because the number of components of the protection and control system is reduced and the number of interfaces is practically reduced to fiber optic cables only. The single exception is the hardwired connections that power the individual devices.

Testing of NCIT Based Protection Applications

The testing of NCIT based protection applications depends on the answer to the following questions:

  • What are we testing?
  • Why are we testing it?
  • How are we going to test it?

Since we already have seen the variety of interfaces and devices that are available when we are using NCITs for protection and control applications, it is important to first answer the question “What are we testing?” If we look at Figure 4, the answers can be that we are testing:

  • An NCIT with a low-power interface
  • An NCIT with an IEC 61850 9-2 interface
  • An NCIT with a low-power interface and a MU
  • A merging unit with a low-power input
  • A protection IED with an IEC 61850 9-2 interface
  • We are testing a protection function or scheme

In case of testing either an NCIT, a merging unit or the combination of both, it is required to apply a true signal reflecting the real operation condition resp. the real direction of information flow. Different to testing of conventional instrument transformers, where the nonlinearity of the devices requires a special care during testing (or the use of alternative methods, e.g. secondary voltage injection), the wide linearity of NCIT over voltage or current allows lower testing signals, if the signal to noise ratio is acceptable. Nevertheless, a statement of linearity, a reference measurement or type test results should be studied before applying a wrong signal magnitude to the NCIT. As stated above, the linearity of NCITs allows the use of test currents lower than rated current multiplied by the over current factors.

This fact is especially helpful for commissioning or other on-site testing as high current sources are typically bulky and heavy. Practice has proven that test currents of several hundred amperes (e.g. 300 A - 800 A) are sufficient for most applications. If test plans or testing guides require higher currents, advanced technologies make even higher test currents available with moderate equipment weight for on-site testing.

Overall, NCIT based systems allow a more cost-effective way of testing which additionally reduces the cost of ownership of the units next to the fact, that explosions and outages are less likely for this technology.

After that, we need to answer “Why are we testing?” Again, we may have a few different answers:

  • It is acceptance testing of the NCIT
  • It is acceptance testing of the merging unit (MU)
  • It is an integration test of the NCIT and the MU
  • It is an integration test of the merging unit and the IED
  • It is a commissioning test of the NCIT
  • It is a commissioning test of the MU
  • It is a commissioning test of the IED

As can be seen from the list above, maintenance testing is not mentioned. The reason is very simple: In IEC 61850 based protection and control systems we can continuously monitor the components of the protection, automation and control, which practically eliminates the need for maintenance testing. Nevertheless, if a deviation of a signal can be observed, and there is no indication of a fault, maintenance tasks should be initiated using proper and accurate testing and diagnostic methods/devices. Thus, a traditional time or condition based maintenance can be substituted with a continuous monitoring plus an event based maintenance where the monitoring part comes for free with the IEC 61850 based system.

Depending on the answer to the two questions we can decide what methods and tools for testing of the components of the protection system can be used.

Testing of NCIT with low-power interface: If NCITs with low-power output should be tested, the required primary current/voltage needs to be injected by the testset and the low-power output must be measured. This allows to determine if the NCIT meets the technical specifications and the requirements of the application.      The measurement error of the NCIT is a systematic, linear deviation in amplitude and phase. Different to conventional units with a ferromagnetic core, this fact allows a) the application of the sensor for a wide current range and application (protection or metering) and b) an easy correction of the amplitude and phase errors by so called correction factors. These correction factors are typically given for every NCIT with low-power output and written on the nameplates. They should be considered during operation and testing. The wide application ranges (i.e. the wide range of primary currents where the NCIT can be used for) make it necessary to deposit the correct ratios in any IED or testset. The question “what are we testing?” is mainly answered. Also, the “how?” has been discussed above and the question “why?” has been explicitly answered as: “we want to commission the NCIT”. But as often, it needs a more detailed answer as well to be able to perform the right test for the right purpose. NCITs have many specifications regarding rated values, thermal capabilities or insulation level. In regards of commissioning testing, the more detailed answer on the “why” is: “we want to ensure correct connection and fulfilled ratio and phase accuracy for the intended purpose in the substation.”

The accuracy of NCITs for rated frequency and rated burden (if applicable) is defined in IEC 61869 as a combination of ratio error ɛ and phase displacement Δφ. For current sensors, index i and for voltage sensors, index u is used.

                                  

                     (1)

Where:

kra is the rated transformation ratio for analog output (index a for analog, index d for digital) that might be multiplied with the correction factor CF

Xp is the r.m.s. value of the input

Ys is the r.m.s. value of the output

The phase displacement Δφ is defined as the phase difference between secondary output and primary input minus phase offset and delay time (both as rated values). To carry out the ratio resp. the accuracy test, we need to inject currents up to several hundred amperes from the test set output or up to several kA using a current booster to the current transformer’s primary side (see Figure 5).

The low power current transformer's secondary side (corresponding voltage signal) is connected to a voltage input of the test device to determine the performance of the NCIT under test. The connectors are specified in IEC 61869-6 and often, the specified RJ 45 connectors are used, connected to the sensor via a pre-packed cable with a certain length (e.g. 5 m).

The following measurements are used to evaluate the performance:

  • Iprim: actual output current at the output that is injected into the low power current transformer’s primary side
  • Vsec: secondary voltage measured and its phase deviation relative to Iprim
  • Ratio: ratio Iprim/Vsec

The ratio can be tested easily with a sufficient accuracy but ratio error and phase displacement measurements up to IEC accuracy class 0.2 (max. ratio error 0.2% and phase displacement max. 10 min) require an accurate current measurement of the actual output current and an accurate voltage and phase measurement at the secondary side of the NCIT.  The linearity of most of the low power instrument transformers opens new dimensions of flexibility and mobility. Especially for on-site and commissioning testing, small powerful and battery operated testsets can be used for this application.

Testing of NCIT with IEC 61850 9-2 LE interface: If we are testing NCITs with IEC 61850 9-2 LE output, we need to be able to apply the required primary current by the testset and monitor the streaming sampled values in order to determine that the NCIT meets the technical specification and the requirements of the application. IEC 61850 defines an Ethernet frame for data transmission. Compared to the previous chapter, we again consider the ratio, and the phase displacement in the context of commissioning testing of the NCIT and MU. Additionally, IEC 61850 based NCIT technologies allow a so called closed loop (commissioning) testing that is presented in the use case below. The accuracy of the digital NCIT is specified similarly to equation 3 but a certain signal handling is necessary. There are different ways to determine the performance of the NCIT, depending on the purpose of the test and the availability of time synchronization between the device and the test object:

  • Testing based on r.m.s. comparison without time synchronization of testset and merging unit (Figure 6a)
  • Testing with time synchronization and comparison of the signals in amplitude and phase (Figure 6b)

Both methods have in common that the primary testset needs to inject a primary current (with or without current booster) and a test computer (either integrated in testset or stand-alone) needs to receive the Sampled Values (SV) stream from the merging unit or the substation network. The MU converts the analog sensor output into a SV stream which is published via the digital interface.

On the receiving side, the test computer (stand-alone or integrated in primary testset) transforms the sampled points to the spectral function of the signal (see figure 8). This Fourier-transformed sampled values signal is filtered with a special Hann window to only retrieve the "signal" at the selected, typically nominal, frequency. This allows frequency-selective measurements to be performed on SV streams and thereby the noise is suppressed. Also, noise can be detected during this test procedure. With this calculated noise-reduced signal, a re-transformation to the discrete signal is performed. At last, the r.m.s. value of this signal is calculated and used for comparison with the reference signal. The expected ratio is 1:1 as the ratio of the NCIT is entered in the MU.

If there is no time synchronization available as in Figure 6a), the performance of the NCIT cannot be evaluated regarding the phase displacement resp. the time delays in the system. The ratio or ratio-error ԑ can be determined as described above. In this case the primary test device injects the sinusoidal test current, the digital SV signal is computed as described above and compared to the “ideal” signal injected by the testset. To test if the polarity is correct, a special saw tooth signal is injected to the NCIT. Even without time synchronization this allows to check the received signal upon correct polarity as positive and negative slopes have different gradients that can be used for evaluation of the polarity.

If phase information is required, the source and the MU need to have the same time base, i.e. they must be time synchronized (see Figure 6b)). The test set generates analog, sinusoidal test signals (here: currents) accurately defined in magnitude and phase, and injects it to the NCIT connected to the time synchronized MU. The phasing of the generated signal is hence known and it can be correlated to the digitized signal having the same time basis. The test computer captures the SV from the merging unit for a detailed assessment of the conversion performance. For a comparison of the generated and measured values, the test computer (time synchronized or use of sample count) subscribes to the SV generated by the test set and uses or displays them as a reference. The test computer can also capture the streamed sampled values and store them for comparison and analysis of the performance of the NCIT. The NCIT performance parameters such as ratio or phase displacement can be calculated accordingly. It is important to connect the test computer close to the physical location of the protection or metering device to test the phase displacement resp. time delay under real operating conditions.

The same methods described above can be used in case of integration testing of an NCIT with a low-power output and a merging unit with a low-power input.

Use case: “Closed-loop testing”: A primary test system as per Figure 6a) can perform closed-loop testing whereby a test signal is injected on the primary side of the current/voltage sensors. The MU converts the sensor output into an SV stream which is published to the substation network. The primary test system then reads the data back from the substation network to perform a variety of different tests regarding NCIT performance, MU performance, wiring and/or setting check. It is thus possible to verify the overall performance in the final substation environment (see Figure 7).

If the whole system including multiple NCITs and MUs is tested using this closed-loop-approach, it is important that the testset can identify the device under test (DUT) among all available NCITs. For this, the primary testset injects a unique saw tooth test signal into the DUT and reads all published SV streams from the substation bus. The testset then searches for the unique signal within all available SV streams and thus identifies the correct channel for testing. Finally, the ratio and polarity are determined as described above.

Protection Functions and Schemes Testing: Once we have verified that the NCITs and MUs are properly connected and provide the expected current and voltage sampled values, we can perform the commissioning testing of protection functions in IEDs or distributed protection schemes by simulating the required fault condition and publishing the IEC 61850 SV streams from the test devices that are part of the test system. In principle, this is similar to injecting secondary currents and voltages to the protection devices for their commissioning testing in conventional hardwired systems instead of injecting fault currents and voltages on the primary side of the conventional instrument transformers.

Biographies

Michael Freiburg is currently working as a product manager at OMICRON electronics in Germany and Austria. He is responsible for instrument transformer test and diagnostic equipment. Before, he has worked as a lecturer and research assistant at the Technical University of Dortmund, Germany. His research interests include design and diagnostics of high voltage equipment and material science.  In his undergraduate studies his focus was on automation and control engineering before he studied power engineering in his post-graduate studies. He received the engineering degree in 2010 and his PhD degree in high voltage engineering in 2014, respectively.

Dr. Alexander Apostolov received his MS degree in Electrical Engineering, MS in Applied Mathematics and Ph.D. from the Technical University in Sofia, Bulgaria. He is Principal Engineer for OMICRON electronics in Los Angeles, CA. He is an IEEE Fellow and Member of the PSRC and Substations C0 Subcommittee. He is past Chairman of the Relay Communications Subcommittee, serves on many IEEE PES WGs. He is a member of IEC TC57 WGs 10, 17, 18, 19, Convenor of CIGRE WG B5.53 and member of several other CIGRE B5 WGs. He is a Distinguished Member of CIGRE. He holds 4 patents and has authored and presented more than 500 technical papers. He is an IEEE Distinguished Lecturer and Adjunct Professor at the Department of Electrical Engineering, Cape Peninsula University of Technology, Cape Town, S. Africa. He is Editor-in-Chief of PAC World Magazine.

Power. Flexible. Easergy.
BeijingSifang June 2016