Automated Testing Of PMU Compliance

Transcription

1 Automated Testing Of PMU Compliance Richard Annell Moe Khorami Murari Mohan Saha ABB AB, Substation Automation Products, Sweden of contact author: Abstract: Validating a Phasor Measurement Unit (PMU) against the IEEE C standard with amendment C a-2014 (from here on referred to as the standard ) poses a challenge in terms of the sheer number of measurements and calculations needed. This paper covers how ABB SA Products AB has accomplished this by designing a test-bench for the RES PMU product. The test-bench is capable of performing settings on the RES being tested, controlling an Omicron GmbH CMC 256plus signal-generator with the CMIRIG-B interface (for applying time-synchronized analogue input stimuli), receiving the C telegrams from the RES under test, calculating Total Vector Error (TVE), Frequency measurement Error (FE) and Rate of change Frequency Error (RFE) and making pass/fail assessments. The test-bench is developed using Keysight VeePRO (formerly known as Agilent VeePRO) which is a graphical programming environment suitable for developing applications which involves control of external instruments and network communication. The test-bench controls the signal-generator using the Omicron GmbH CMEngine interface which is a software interface which enables controlling Omicron GMBH equipment using third party software and communicates with the RES using TCP/IP. Index Terms: PMU, Synchrophasors, IEEE Std C standard, IEEE Std C a amendment, IEEE C guide, Automated Testing, IEEE Std C standard compliance 1. INTRODUCTION The introduction of PMU s to the power industry in the early 1990s helped revolutionize the way we analyze and control power systems. All the measurements taken are synchronized and time-stamped by a precision timing device. Wide Area Systems are implemented allowing the networks to operate closer to their capacity while maintaining the system security [1]. Wide Area Monitoring, Protection and Control System is a response to the "Smart Grid" concept on transmission and sub-transmission level in order to increase network operator's awareness of grid performance and security while making our power system smarter and more reliable. Accurate Phasor Measurement Units, fast and reliable communication infrastructure as well as smart wide area applications are the building blocks of wide area systems [2]. However verification of the PMU s accuracy requires special test procedures, time sources and very accurate testing instrument [1]. The accuracy of synchrophasors (amplitude and angle) is expressed by a quantity called Total Vector Error (TVE). According to the standard, TVE is an expression of the difference between a perfect sample of a theoretical synchrophasor and the estimate given by the unit under test at the same instant of time. The value is normalized and expressed as per unit of the theoretical phasor [3]. The PMU subjected to the described automated testing procedure in this paper is the ABB RES During development of RES , investigations were made regarding the possibility to have an external facility validate these requirements. However, at that point in time, this was not a feasible option because there was no external laboratory available that could certify the PMU s compliancy with the standard. Considering the amount and the complexity of the measurements that has to be performed in order to validate compliance with the standard, there is a lot to be gained by automating these tasks to the largest extent possible. For this reason, ABB SA Products AB has developed an automated testbench called RES_ComplianceTests_C (from here on referred to as the test-bench ) which significantly facilitates the testing procedure. The test-bench is intended for internal ABB use only. The paper continues with describing the hardware and software aspects of this test-bench.

2 2. TECHNICAL SOLUTION, HARDWARE A basic requirement for performing these measurements is the possibility to generate analogue input stimuli to the PMU which is synchronized to an absolute reference in time. Another basic requirement is that, in addition to the above, the generated analogue input stimuli have to be generated with a sufficiently high accuracy. The signal-generator from Omicron GMBH, model CMC 256plus (from here on referred to as the signal-generator ), used in this solution is capable of this when used in conjunction with the external IRIG-B time synchronization module. Fig. 1. Schematic diagram of the test setup Although it is possible to have the signal-generator act as a time reference in this application, an external station clock was selected for synchronizing both the signal generation and the RES PMU under test. Connecting an oscilloscope with one channel connected to the PPS (Pulse Per Second) output of the station clock and one channel connected to the first voltage output of the signal-generator during the measurements shows that it is possible to achieve a very high accuracy in terms of phase-shift in reference to PPS using this hardware setup. The technical specifications for the Omicron GMBH CMIRIG-B module states that the time error of time reference source to analogue outputs is typically better than 1µs with a maximum error of 5µs (for CMC 256plus) [4]. Measurements performed as described above showed that the voltage outputs stayed well within the typical limit of 1µs (with the point in time when the positive zero crossing of the output voltage passes through zero volts defined as zero degrees). According to the IEEE C guide, general purpose test equipment that can be used to test PMU functions should be 4 to 10 times more accurate than the test tolerance, but this will vary depending on the circumstance. This means a Test Uncertainty Ratio (TUR) equal to 4 to 10. TUR is a quantity which describes how much better the test equipment is than the equipment being tested [5]. Since the test-bench sets parameters for the signal generation via the CMEngine interface directly on the signal-generator, technical specifications for the signal-generator can be used for determining the accuracy of the generated signals [6]. The technical specifications for the signal-generator specify magnitude accuracy for the current outputs to be better than 0.04% of setting plus 0.01% of range (ranges are 1.25A and 12.5A). As an example of this, at the magnitude of 1A, the magnitude error is less than A (0.0004A A), which equals %. Phase-shift error is specified to be better than 0.02 (degrees). Even if adding a phase-shift error corresponding to 1µs (0.018 at

3 50Hz) to account for timing inaccuracy using the external IRIG-B timing module, the phase-shift error is still better than In terms of TVE, the % amplitude error and angle error together will represent an error of less than 0.085% in the generated signal. TUR in this case when validating against the 1% TVE limit commonly specified in the standard is higher than At the set magnitude of 0.1A (10% of 1A), the TUR still exceeds 5.6. For voltage outputs, the technical specifications for the signal-generator specify magnitude accuracy for the voltage outputs to be better than 0.04% of setting plus 0.01% of range (ranges are 150V and 300V). As an example of this, at the magnitude of 110V, the magnitude error is less than 0.059V (0.044V+0.015V), which equals 0.054%. Phase-shift error is specified to be better than Even if adding a phase-shift error corresponding to 1µs (0.018 at 50Hz) to account for timing inaccuracy using the external IRIG-B timing module, the phase-shift error is still better than In terms of TVE, this equates to an error of less than 0.086% in the generated signal. The TUR in this case when validating against the 1% TVE limit commonly specified in the standard is higher than At the set magnitude of 11V (10% of 110V), the TUR still exceeds 5.3. Using a separate signal source such as a plug in board for a PC combined with an external current amplifier would most likely present a greater challenge in terms of achieving such a high TUR value. 3. TECHNICAL SOLUTION, SOFTWARE From a software point of view, the main requirements on the test-bench performing these measurements are the ability to: Perform settings on the PMU regarding rated frequency, performance class and report rate Calculate and apply analogue input stimuli based on settings made and test at hand Receive C telegrams from the PMU under test Calculate TVE, FE and RFE Calculate delay-time, response-time and over-/under-shoot when applicable Calculate limits for the above based on current PMU settings and test at hand Perform PASS/FAIL assessment Logging of retrieved and calculated data, limits and results (for post-analysis) If the above requirements are met and combined with the ability to loop through the settings and test types required, a high degree of automation is achieved. Using the Omicron GMBH software option CMEngine, which enables controlling the signal-generator from any third party software that is capable of utilizing ActiveX automation, it is possible to develop an application that combines functionality like performing settings on the PMU with applying analogue input stimuli and collecting data frames from the PMU under test. Keysight VeePRO constitutes a development environment which is suitable for this and was chosen for the development of the testbench [7]. The test-bench is capable of validating the requirements specified in the following sections of the standard for one voltage phasor and one current phasor simultaneously: Section "Steady-state compliance" Section "Dynamic compliance-measurement bandwidth" Section "Dynamic compliance-performance during ramp of system frequency" Section "Dynamic compliance-performance under step changes in phase and magnitude" The following 10 test types were defined in the test-bench in order to cover the requirements specified above: MagnTest (section 5.5.5, Table 3, Influencing quantity: Signal magnitude) PhaseTest (section 5.5.5, Table 3, Influencing quantity: Phase angle) FreqTest (section 5.5.5, Table 3 & 4, Influencing quantity: Signal frequency) InterfTest (section 5.5.5, Table 3 & 4, Influencing quantity: Out-of-band interference) HarmTest (section 5.5.5, Table 3 & 4, Influencing quantity: Harmonic distortion) ModMagn (section 5.5.6, Table 5 & 6, Amplitude modulation) ModPhase (section 5.5.6, Table 5 & 6, Phase modulation) FreqRamp (section 5.5.7, Table 7 & 8, Linear frequency ramp) AmplStep (section 5.5.8, Table 9 & 10, Amplitude step)

4 PhaseStep (section 5.5.8, Table 9 & 10, Phase step) In the standard, table 3, 5, & 7 specifies TVE limits. Table 4, 6, & 8 specifies FE and RFE limits. Table 9 specifies phasor performance during step change in terms of response time, delay time and over-/under-shoot. Table 10 specifies frequency and rate of change frequency performance during step change in terms of response time [3]. Table 1 shows the supported settings in the RES in terms of rated frequencies (Fr), Performance Classes (PC) and Report Rates (RR): Fr PC RR Fr PC RR P M P M Table 1. Supported settings, RES The settings in Table 1 constitute 26 setting combinations to be tested. Combined with the 10 defined test types in the test-bench, 260 different test cases have to be executed. Testing hardware with another rated input current for example, requires all these 260 test cases to be repeated. Combined with the fact that every test performed requires a substantial amount of data to be collected and analyzed, the need for automation of the task at hand becomes obvious. 3.1 TEST-BENCH, MAIN WINDOW Figure 2 below shows an example of the main window of the test-bench during execution.

5 Fig. 2. Main window of the test-bench In the test-bench, IP-address of the PMU, rated current, rated voltage and the phasor numbers for the voltage and the current phasor is entered. The operator can choose which rated frequencies to test at (50 Hz, 60 Hz or both), which performance classes to test (P, M or both) and which report rates to test for each rated frequency. The operator can also choose whether to execute one test type or all of the supported test types in the application. Figure 3 below shows examples of selecting these parameters. Fig. 3. Selecting rated frequency, performance class and report rate for Fr=50 Hz Additionally, there are individual settings available for most of the test types. These settings are intended to enable executing only a specific test type with a reduced set of influencing quantities resulting in a much shorter execution time (for experimental purposes). 3.2 AUTOMATIC SELECTION OF APPLIED ENERGIZING QUANTITIES The standard may specify different ranges of applied energizing quantities depending on for example performance class used. Figure 4 below shows an excerpt from the standard (section 5.5.5, part of table 3), specifying ranges of applied influencing quantities and TVE limits. Notice the differences in

6 the Range -column between P and M performance class for signal frequency and signal magnitude for voltage. Fig. 4., Excerpt from the standard section 5.5.5, part of table 3 Since the test-bench is aware of the settings being applied during testing, the test-bench is also capable of automatically selecting the correct range of influencing quantities to use for the test at hand.

7 Figure 5 below shows example code for selecting the voltage range to be used for the test type MagnTest. Fig. 5. Code snippet, selection of voltage magnitudes Similar methodology is used for all test-types defined in the test-bench when applicable. For example when selecting the amount of harmonic distortion to apply depending on performance class, when selecting which interfering frequencies to apply and the magnitude of them and when selecting the start and stop frequencies during frequency ramp testing etc. 3.3 AUTOMATIC SELECTION OF LIMITS The standard may also specify different limits depending on for example performance class used. Figure 6 below shows an excerpt from the standard (section 5.5.5, table 4), specifying ranges of applied influencing quantities and FE and RFE limits. Notice that there are several differences in the Error requirements -column between P and M performance class. Additionally, there may be limits that are not applicable or suspended. Such as the RFE limits for harmonic distortion test and out-ofband interference test in the figure below. The test-bench is capable of detecting when this is the case and sets these limits to NA (not applicable) for such tests. However, FE and RFE are still measured and logged in those cases. Fig. 6. Paragraph 5.5.5, table 4 from the standard

8 Again, since the test-bench is aware of the settings being applied during testing, the test-bench is also capable of automatically selecting the correct limits to use for the test at hand. Figure 7 below shows example code for selecting the RFE limit to be used. Fig. 7. Code snippet, selection of RFE limit Figure 8 below shows example code for selecting the limits and the validity of the limits when testing with harmonic distortion as the influencing quantity. Notice that the variable RFELimitValid is set to 0 (false) if the performance class is M, and that the FE limit is set depending on the set report rate if performance class is M. Fig. 8. Code snippet, selection of limits and validity of limits

9 3.4 STEADY-STATE TEST-TYPES The first five test-types defined in the test-bench which covers section in the standard, Steadystate compliance, uses the predefined signal quantities available via the CMEngine interface for the signal-generator in order to define the analogue input stimuli for the test at hand. For these test-types there is no need to calculate and download waveforms to the signal-generator MAGNTEST Test type MagnTest sets frequency, phase shifts and magnitudes for the voltage and current outputs used on the signal-generator and applies them synchronized to the time signal from the station clock. The only quantity that varies during the test is the magnitude of the applied signals. For every magnitude applied, C telegrams are received from the PMU under test, TVE, FE and RFE calculated and compared to limits when applicable. Figure 9 below shows an example of this test during execution. Notice the PASS/FAIL indicators on the far right showing NA when the test-bench has determined that a limit is not to be applied (not specified in the standard or suspended). However, as can be seen in figure 9 below, the test-bench will still measure those quantities. Fig. 9. Example of MagnTest in progress

10 3.4.2 PHASETEST Test type PhaseTest sets frequency, phase shifts and magnitudes for the voltage and current outputs used on the signal-generator and applies them synchronized to the time signal from the station clock. The only quantity that varies during the test is the phase-shift of the applied signals. For every phase-shift applied, C telegrams are received from the PMU under test, TVE, FE and RFE calculated and compared to limits when applicable. Figure 10 below shows an example of this test during execution. Fig. 10. Example of TVE measurements for the voltage phasor during PhaseTest FREQTEST Test type FreqTest sets frequency, phase shifts and magnitudes for the voltage and current outputs used on the signal-generator and applies them synchronized to the time signal from the station clock. Care has to be taken in this case when applying frequencies that are not evenly dividable. For example, if applying a frequency of 50.1 Hz, time synchronization is only possible every 10 Th second in order to keep the signal uninterrupted in terms of avoiding sudden phase shifts. The only quantity that varies during the test is the frequency of the applied signals. For every frequency applied, C telegrams are received from the PMU under test, TVE, FE and RFE calculated and compared to limits when applicable. Figure 11 below shows an example of this test during execution. Fig. 11. Example of TVE measurements for the voltage phasor during FreqTest INTERFTEST Test type InterfTest sets frequency, phase shifts and magnitudes for the voltage and current outputs used on the signal-generator and applies them synchronized to the time signal from the station clock. In this test-type, both the available signal definition memories per channel in the signal-generator are used. One for the applied fundamental frequency and the other one for the applied interfering frequency. Care has to be taken in this case when applying interfering frequencies that are not evenly dividable. For example, if applying an interfering frequency of 50.1 Hz, time synchronization is only possible every 10 Th second in order to keep the signal uninterrupted in terms of avoiding sudden phase shifts. The only quantity that varies during the test is the interfering frequency added to the applied signals. For every interfering frequency applied, C telegrams are received from the PMU under test, TVE, FE and RFE calculated and compared to limits when applicable.

11 Figure 12 below shows an example of this test during execution. Fig. 12. Example of TVE measurements for the voltage phasor during InterfTest HARMTEST Test type HarmTest sets frequency, phase shifts and magnitudes for both the fundamental signal and the harmonic applied for the voltage and current outputs used on the signal-generator and apply them synchronized to the time signal from the station clock. The only quantity that varies during the test is the number of the harmonic applied. For every harmonic applied, C telegrams are received from the PMU under test, TVE, FE and RFE calculated and compared to limits when applicable. Figure 13 below shows an example of this test during execution. Fig. 13. Example of TVE measurements for the voltage phasor during HarmTest 3.5 DYNAMIC TEST-TYPES The five remaining test-types defined in the test-bench which covers section 5.5.6, and in the standard, Dynamic compliance, calculates waveforms dynamically depending on the PMU settings in use and the test at hand. The calculated waveforms are downloaded to the signalgenerator via the CMEngine interface in order to define the analogue input stimuli. The magnitude of the outputs and the playback rate is set and the playback of the downloaded waveforms is started synchronized to the time signal from the station clock MODMAGN Test type ModMagn calculates the waveforms used for modulated analogue input stimuli using basically the same mathematical algorithms as defined in the standard in paragraph Figure 14 below shows an example of the waveform calculation used for modulated analogue input stimuli. During modulation of magnitude, the phase angle modulation is set to zero. Fig. 14. Code snippet, example of calculation of modulated waveforms

12 Where: Xm=amplitude, Kx=amplitude modulation factor, Ka=phase angle modulation in degrees, df=modulation frequency in Hz, f=fundamental frequency in Hz and t=time in s. A, B and C are containers for the result of the calculations for phase A (L1), B (L2) and C (L3) respectively. For comparison, figure 15 below shows an excerpt from the standard, section 5.5.6, mathematically describing the waveforms during modulation. The only tangible difference is the usage of radians rather than degrees. Fig. 15. Excerpt from the standard, section 5.5.6, mathematical representation of the waveforms during modulation The calculated waveforms are downloaded to the signal-generator. Playback rate and magnitudes for the voltage and current outputs used on the signal-generator is set and playback is started synchronized to the time signal from the station clock. The only quantity that varies during the test is the modulation frequency applied. For every modulation frequency applied, C telegrams are received from the PMU under test, TVE, FE and RFE calculated and compared to limits when applicable. Figure 16 below shows an example of this test during execution. Fig. 16. Example of TVE measurements for the voltage phasor during ModMagn MODPHASE Test type ModPhase calculates the waveforms used for modulated analogue input stimuli using the same mathematical algorithms as test type ModMagn. During modulation of phase angle, the amplitude modulation factor is kept to zero. The calculated waveforms are downloaded to the signal-generator. Playback rate and magnitudes for the voltage and current outputs used on the signal-generator is set and playback is started synchronized to the time signal from the station clock. The only quantity that varies during the test is the modulation frequency applied. For every modulation frequency applied, C telegrams are received from the PMU under test, TVE, FE and RFE calculated and compared to limits.

13 Figure 17 below shows an example of this test during execution. Fig. 17. Example of TVE measurements for the voltage phasor during ModPhase FREQRAMP Test type FreqRamp calculates the waveforms used for analogue input stimuli with a ramping frequency. The calculation is divided into three parts; waveform before ramping of the fundamental frequency takes place, waveform during ramping of the fundamental frequency and waveform after ramping of the fundamental frequency. Ramp rate and ramp range is set automatically according to the standard, paragraph 5.5.7, table 7. The calculated waveforms are downloaded to the signal-generator. Playback rate and magnitudes for the voltage and current outputs used on the signal-generator is set and playback is started synchronized to the time signal from the station clock. The only quantity that varies during the test is the direction of the frequency ramp applied (positive or negative). C telegrams are received from the PMU under test, exclusion intervals is automatically calculated and applied, TVE, FE and RFE calculated and compared to limits. Figure 18 below shows an example of this test during calculation of TVE, FE and RFE. Fig. 18. Example of FreqRamp in progress during calculation of TVE, FE and RFE

14 3.5.4 AMPLSTEP Test type AmplStep calculates the waveforms used for analogue input stimuli with a step change in amplitude. The calculation is divided into two parts; waveform before step change of the applied amplitude and waveform after step change of the applied amplitude. In order to achieve the required accuracy of at least one-tenth of the used report rate in response- and delay-time measurement, the whole process is repeated ten times. Each time, the step-change is delayed one-tenth of the time corresponding to the selected report rate. For example, if the report rate 10 is used, each iteration delays the step-change 10ms compared to the previous iteration. The received C telegrams from the PMU are then correlated by subtracting the time stamps with the delay used in the current iteration and finally interleaved, thus creating a set of data with 10 times higher resolution than the used report rate would provide in itself. The calculated waveforms are downloaded to the signal-generator. Playback rate and magnitudes for the voltage and current outputs used on the signal-generator is set and playback is started synchronized to the time signal from the station clock. The only quantity that varies during the test is the direction of the amplitude step applied (positive or negative). C telegrams are received from the PMU under test, response and delay times, over-and under-shoot are determined and compared to limits. After the final iteration, the final assessment is made. Figure 19 below shows an example of this test after the final iteration. Fig. 19. Example of AmplStep in progress during assessment of results

15 3.5.5 PHASESTEP Test type PhaseStep calculates the waveforms used for analogue input stimuli with a step change in phase shift. The calculation is divided into two parts; waveform before step change of the applied phase shift and waveform after step change of the applied phase shift. In order to achieve the required accuracy of at least one-tenth of the used report rate in response- and delay-time measurements, the whole process is repeated ten times. Each time, the step-change is delayed onetenth of the time corresponding to the selected report rate. For example, if the report rate 10 is used, each iteration delays the step-change 10 ms compared to the previous iteration. The received C telegrams from the PMU are then correlated by subtracting the time stamps with the delay used in the current iteration and finally interleaved, thus creating a set of data with 10 times better granularity than the used report rate would provide if the methodology described above wasn t used. The calculated waveforms are downloaded to the signal-generator. Playback rate and magnitudes for the voltage and current outputs used on the signal-generator is set and playback is started synchronized to the time signal from the station clock. The only quantity that varies during the test is the direction of the amplitude step applied (positive or negative). C telegrams are received from the PMU under test, response and delay times, over-and under-shoot are determined and compared to limits. After the final iteration, the final assessment is made. Figure 20 below shows an example of this test after the final iteration. Fig. 20. Example of PhaseStep in progress during assessment of results

16 3.6 DATA LOGGING The test-bench logs the results in multiple ways providing the possibility for different levels of post analysis should so be desired. The first level is a simple summary of all tests that has been executed during a session including information about test type, rated frequency, performance class and report rate and whether the test at hand was assessed as passed or failed. This log can be viewed directly from within the test-bench after concluding a test session. Figure 21 below shows an example of this. Fig. 21. Example of viewing the summary log The test-bench automatically generates screen shots of the complete operators interface after concluding each test. These screen shots are automatically named depending on the test type and settings used including a user defined prefix. For example, the filename 110V1AMPS_Fr50_P_PhaseTest_Rr10.png is a screen shot of the test type PhaseTest with the user defined prefix 110V1AMPS, executed with the rated frequency at 50 Hz, performance class set to P, report rate set to 10. Figure 22 below shows an example of this. Fig. 22. Example of automatically generated screen shot during execution

17 Additionally, a comprehensive log using the Keysight VeePRO [7] proprietary dataset format is generated automatically for each test executed. This log contains both calculated values in terms of TVE, FE, RFE, measured times and the raw telegrams received from the PMU during each test. This log also contains the selected limits during the test and whether that limit is applicable or not for the test at hand. The log is in text format and in a machine readable form making it easy to post process. This enables more complex post analysis to be performed. Figure 23 below shows a partial example of this. Fig. 23. Partial example of automatically generated log during execution 4. DEGREE OF AUTOMATION When using this test-bench during compliance testing of the RES , the degree of automation achieved is to be considered very high. If executing what can be referred to as a full test session, meaning that both rated system frequencies, both performance classes, all supported report rates and test types are included, the total unattended execution time can exceed 55 hours. During that time, more than 3.8 million C telegrams will be received from the PMU under test, more than 16 million calculations in terms of TVE, FE, RFE, response time, delay time, overshoot and under-shoot will have been performed. PASS/FAIL assessments for all permutations of selected settings and testtypes and logging of all data needed for post analysis and report creation performed. Or, to put it another way, in less than three days, all supported test types will have been executed for both rated system frequencies, both performance classes, all supported report rates and test types. Additionally, the repeatability achieved using this test-bench combined with the degree of automation achieved makes it very useful for regression testing. Figure 24 below shows an example of statistics shown in the operators interface after completion of a test session. Fig. 24. Example of statistics shown in the operator s interface of the test-bench

18 5. CONCLUSIONS Testing a PMU in regards to compliance with the standard requires a great extent of automation. By developing and using this test-bench, ABB SA Products AB has achieved a high degree of automation for compliance testing of the RES PMU. This was achieved by including and combining settings functionality, generation functionality, data collection functionality, result assessment functionality and data logging functionality in the same testbench. By eliminating solutions such as separate signal sources feeding external amplifiers, high accuracy in terms of generated analogue input stimuli was achieved without the need to and the complexity involved in measuring the generated analogue input stimuli with external instruments or comparing the PMU measurements with a reference PMU measuring the same analogue input stimuli. 6. REFERENCES [1] M. Khorami, Phasor Measurement Units (PMU) Applications in Power Systems (EuroDoble Colloquium, Barcelona, October 2013, Paper presentation D-4) [2] M. Khorami, Real Time Application of Synchrophasors for Improved Reliability (EuroDoble Colloquium, Barcelona, October 2013, Paper presentation A-3) [3] IEEE Standard for Synchrophasors Measurements for Power Systems, IEEE Std C (Revision of IEEE Std C ) with amendment IEEE Std C a [4] Omicron GMBH CMIRIG-B, Technical Data: (accessed ) [5] IEEE Guide for Synchronization, Calibration, Testing, and Installation of Phasor Measurement Units (PMUs) for Power System Protection and Control, IEEE Std C [6] Omicron GMBH CMC 256plus, Technical Data: (accessed ) [7] Keysight VEE, development environment: &cc=US&lc=eng (accessed )