Parallel-Connected Voltage Generators for the Characterization of MV VTs up to 150 kHz

Parallel-Connected Voltage Generators for the Characterization of MV VTs up to 150 kHz Mohamed Agazar Electrical Metrology Division Laboratoire National de Métrologie et d’essais Trappes, France [email protected] Claudio Iodice Dept. of Engineering University of Campania “Luigi Vanvitelli” Aversa (CE), Italy [email protected] Antonio Delle Femine Dept. of Engineering University of Campania “Luigi Vanvitelli” Aversa (CE), Italy [email protected] Mario Luiso Dept. of Engineering University of Campania “Luigi Vanvitelli” Aversa (CE), Italy [email protected] Daniele Gallo Dept. of Engineering University of Campania “Luigi Vanvitelli” Aversa (CE), Italy [email protected] Abstract—Recent advance in semiconductor technology has enabled the development of power converters with higher and higher switching frequencies. Some emerging technologies allow power converters to be directly connected to Medium Voltage (MV) power grids. Consequently, high-frequency components—specifically, harmonics related to the switching frequency—will be present within MV networks. To ensure the reliable and stable operation of these grids, it is essential that such disturbances are accurately measured. This scenario enhanced performance requirements on Instrument Transformers (ITs), particularly in terms of measurement accuracy over an extended frequency range. In order to properly assess ITs’ frequency behaviour, it is crucial to implement a generation architecture, which allows to generate reduced-amplitude spectral components up to hundreds of kilohertz superimposed to the power frequency component. Such an architecture is not currently available in the market and that is a huge gap between the actual state of the art and the metrological needs which have been emerging in recent years. In this paper, a generation setup for the characterization of Voltage Transformers (VTs) in the frequency range including power frequency (50/60 Hz) and spectral components within the range 9 kHz - 150 kHz is provided. The proposed architecture consists of two grounded and parallel-connected voltage generators to separate the generation of the power frequency component from the generation of high-frequency tones. Experimental results related to the working operations of the proposed architecture are discussed. Keywords—Instrument Transformer, Voltage Transformer, Power System Measurements, High-Frequency Components, Medium Voltage I. INTRODUCTION The widespread adoption of switching devices such as inverters, bulky power electronic converters and active filters, used both as loads and as components within generators, particularly in renewable energy systems, has led to a significant increase in conducted electromagnetic disturbances on grid voltage and current. These disturbances, observable even at the Medium Voltage (MV) level, can extend up to several hundred of kilohertz as a result of harmonics related to the switching frequencies. Such high-frequency conducted emissions can interfere with Power Line Communication (PLC), a communication technology extensively employed in power grids to support essential functions, including monitoring, control and automation [1]. This interference degrades the quality and integrity of transmitted information and possible failures of vital grid operations can occur [2]-[4]. High-frequency emissions originating from a single device may couple with other devices, disrupting their control systems and possibly leading to malfunctions or complete operational failures. Additionally, these emissions increase the Root Mean Square (RMS) current, which, along with an intensified skin effect, results in high temperatures and, consequently, accelerates aging, and reduces the useful life of components. Modern power converters, such as Direct Current/Direct Current (DC/DC), Alternating Current/Direct Current (AC/DC), and AC/AC converters, use new generation switching components like Silicon Carbide (SiC) MetalOxide-Semiconductor Field-Effect Transistors (MOSFETs) [5], [6]. These technologies enable higher switching frequencies and ease straight connection to MV networks. As a result, nowadays MV grids face measurement challenges similar to those already faced in Low Voltage (LV) systems [7]–[10]. These changes need enhanced performance from measurement systems, especially in terms of accuracy and bandwidth, to ensure reliable operation. Consequently, Instrument Transformers (ITs), the primary elements in nearly all voltage and current measurement chains, must match these evolving requirements. From a standardization perspective, the existing standards governing IT performance at power frequency (50/60 Hz) are IEC 61869 parts 6 [11], 14 [12], and 15 [13]. These documents address Low Power Instrument Transformers (LPITs) for MV and High Voltage (HV) networks and define accuracy requirements up to 20 kHz. The second revision of IEC 61869-1 (2024) [14] now extends accuracy requirements to frequencies up to 150 kHz and even 500 kHz. However, these standards do not provide guidance on measurement methodologies, generators architectures, test procedures, reference instrumentation, nor uncertainty evaluation for frequencies beyond the power frequency range [15]. As a result, manufacturers and calibration laboratories retain full discretion in assessing compliance with accuracy specifications above the power frequency, thereby limiting end-users’ ability to independently verify product performance. The activity here presented is developed according to the framework of the European Partnership in Metrology (EPM) 22NRM06 “ADMIT” [16], which aims to establish suitable parameters for the definition of the accuracy of Voltage and Current Transformers (VTs and CTs), and the related measurement instrumentation, in the frequency range up to 150 kHz. A non-exhaustive list of the activities that will be carried out within this project is: definition of realistic waveforms, accuracy parameters and related procedures for the performance evaluation of ITs up to 150 kHz, development of new systems for the generation of AC (50/60 Hz) or DC voltages (up to 36 kV) and currents (up to 2 kA) with spectral components of reduced amplitudes up to 150 kHz, development of reference MV voltage and current sensors up to 150 kHz, contribution to the development of new standards about accuracy evaluation of MV ITs up to 150 kHz, within the International Electrotechnical Commission (IEC) Technical Committee (TC) 38 “Instrument Transformers”. This paper deals with the design and implementation of a generation and measurement setup that allows for characterizing VTs up to 35 kV/50 Hz, in a frequency range from 9 kHz to 150 kHz. The structure of the paper is as follows. Section II provides a literature review and Section III describes the implementation of the proposed setup. Section IV deals with some preliminary results and Section V draws the conclusions. II. LITERATURE REVIEW A quite wide literature about the measurement of the frequency behaviour of VTs and CTs beyond the power frequency exists. For brevity, here only some papers are cited [15], [16]. It is well recognized within the scientific community that assessing the frequency behaviour of instrument transformers (ITs) requires test signals consisting of a fundamental component (either DC or 50/60 Hz AC), combined with additional spectral components of lower amplitude [17]. However, to the best of the authors’ knowledge, the frequency range investigated in existing studies is typically limited to 10 kHz [18]-[20]. This limitation is likely due to technological constraints, as neither commercial solutions nor custom implementations capable of generating and measuring such composite waveforms beyond a few tens of kilohertz are currently available. Another field of application which has some common points with the research topic of this paper is related to the measurement of high-voltage pulses [21]. Common points between the two fields include the voltage levels (up to tens of kilovolts or higher) and the frequency range (from DC up to several hundred kilohertz). Nonetheless, significant differences exist that preclude the direct application of pulse measurement techniques to the frequency characterization of ITs designed for power system applications up to 150 kHz. Specifically, the nature of the test signals is greatly different: high-voltage pulse measurements typically involve shortduration square or exponential pulses (on the order of 0.1 to 10 µs), whereas ITs frequency characterization deals with waveforms consisting of a fundamental component and superimposed high-frequency spectral elements. A generation architecture different from the one proposed in this paper has been designed and implemented in the recent years [24]. It consists in separating the generation of the power frequency component from the high-frequency components, by using two different generators series-connected. A Step-Up Transformer (SUT) is fed by a Low-Frequency Amplifier (LFA) to generate the fundamental component at MV level. A High-Frequency Amplifier (HFA) is connected in series with the secondary winding of the transformer to generate harmonic components up to 150 kHz. The series-connection of LFA/SUT and HFA allows for the synthesis of a HV signal containing both the fundamental frequency (50 Hz) and HF harmonics. Two reference devices are used for measurement, one for fundamental tone and one for the HF tones. The biggest issues of this architecture are related to the generators grounding and to the high-frequency voltage drop across the secondary side of the SUT, which is solved by placing a capacitor parallel-connected to the secondary windings of the SUT itself [25]. In [25], for hardware limitations, the maximum voltage of this setup is set to 3.6 kV. III. MEASUREMENT SETUP A. Setup architecture The generation architecture here proposed consists of the parallel connection between the Power Frequency Component (at 50/60 Hz) generator and the High-Frequency Components (HFCs) generator. The basic block diagram of the proposed solution is shown in Fig. 1. It proposes a classical characterization method based on the comparison [17]-[19] of the outputs of a reference VT and of a VT under test, by means of a comparator. Both the comparator as well as the reference VT should have sufficient accuracy in the frequency range from the power frequency up to 150 kHz. Fig. 1. Block scheme of the proposed generation architecture Power Frequency Generator High Frequency Generator + Reference VT VT under test Comparator The Power Frequency Generator (PFG) is obtained by means of an Arbitrary Waveform Generator (AWG), cascaded by a low-voltage amplifier, cascaded, in turns, by a step-up transformer to reach the required voltage levels. Instead, the high-frequency generator is obtained by means of an Arbitrary Waveform Generator (AWG), cascaded by another lowvoltage amplifier, having an output frequency range from DC up to 150 kHz. The proposed measurement setup is shown in Fig. 2. As it can be seen from Fig. 2, the sum of the PFC and of the HFC is simply obtained by the parallel connection of the output of the Step-Up Transformer (SUT) and of the HighFrequency Amplifier (HFA). B. Setup implementation The applied voltage signals are provided by a twochannels arbitrary waveform generator (AWG) Keysight 33612A (± 10 V output range, 660 MHz maximum sampling rate, and 64 million of samples per channel of onboard memory). The comparator is realized by a Data Acquisition System (DAQ) PicoScope 5244D MSO (16-bit resolution, 200 MHz bandwidth, ± 20 V maximum input voltage), which acquires the primary and secondary waveforms of the VT under test. The output of the first channel of the AWG (AWG ch. 1) is connected to a voltage amplifier (± 120 V, ± 50 A, 6 kVA) feeding the VT under test through a 100 V / 35 kV SUT with a power capacity of 5 kVA. The output of the second AWG (AWG ch. 2) is connected to a power amplifier N4L LPA400 (± 400 V, 50 mA, DC – 1 MHz, selectable gain). In the case at hand, the device under test (DUT) was represented by a divider designed and implemented within a collaboration between LNE and University of Campania “Luigi Vanvitelli”; it has a scale factor of 500:1 and uncertainties on ratio error of 200 μV/V (level of confidence 95%) from DC up to 1 MHz, and it is an RC divider (100 MΩ/100 pF). Such a divider has been placed in the DUT slot of the setup, and it was used to measure the output voltage generated by the whole setup (i.e., the combination of the outputs of the two generators). From this point on, this divider will be addressed as Universal Divider (UD). Two other dividers have been used to measure 50 Hz and high-frequency voltages respectively, just at the output of the corresponding generators. From this point on, these two dividers will be addressed as Divider n.1 (D1) and Divider (D2), respectively. Divider D1 is a commercial RC divider (300 MΩ/10 pF, ratio 1000:1, accuracy 0.1% at 50 Hz and 1% at 150 kHz). Divider D2 is a commercial RC divider (100 MΩ/3 pF, ratio 1000:1, accuracy 0.1% at 50 Hz and 1% at 150 kHz). Moreover, a full remote control of the setup has been implemented in LabVIEW environment. The realized software allows to choose generation parameters, to enable and disable the generation, and to monitor overvoltage and overcurrent situations. C. Blocking components In this generation architecture, it is necessary to protect each generator from the current coming from the other generator. Therefore, two blocking filters have been designed and included. First, a resistor R1 and a capacitor C1 (rated at 50 kV too) are used to protect the step-up transformer from highfrequency current. C1 is a polypropylene film capacitor with voltage coefficient at 50 Hz < 0.2% and linearity error up to 150 kHz < 0.1%. These elements make up the blocking element n. 1 (see Fig. 2) and create a low-pass effect with a cutoff frequency equal to  , ≅636 , ensuring that HF components are sufficiently attenuated on the transformer side. Moreover, it should be avoided that an amount of 50 Hz voltage drops across the resistor R1. Therefore, DUT impedance at 50 Hz should be much larger than R1 resistor. For the values chosen: |  50 |  ∙∙  ⁄ " # $ #%∙&'#( # 30.33 *Ω (1) |  50 |≫-1100 /Ω (2) Additionally, a HV capacitor, C2 (rated at 50 kV), is placed on the high-frequency amplifier side. C2 is a ceramic capacitor (N4700 dielectric). Together with the output impedance R2 (50 Ω) of the amplifier itself, it makes up the blocking element n. 2 (see Fig. 2) and creates a high-pass effect with a cutoff frequency equal to  , ≅1.5 *, ensuring that the 50 Hz component is almost totally removed. Moreover, as it was previously stated for blocking element n. 1, it should be avoided that an amount of high-frequency voltage drops across capacitor C2. As DUT is made up of resistors and capacitors parallel-connected, for high frequency DUT capacitors dominate. Therefore, for the case at hand, the Fig. 2. Setup implementation of the proposed generation setup influence of DUT resistive part is negligible in high frequency, and it can be stated that DUT capacitance should be much smaller than capacitor C2. For the values chosen: 0123435367 100 89≪;22 =9 (3) Following the calculation provided in (3), only  4>  6>    5% of the generated high-frequency voltage drops across capacitors C2, which is acceptable. Obviously, if DUT is changed, it is necessary to rescale blocking elements as well. Blocking elements are not influenced by dividers D1 and D2. Divider D1 has a negligible effect on blocking element n. 1, because ;≫; . On the other side, divider D2 does not impact the behaviour of blocking element n. 2, whatever the value of ;. Table I provides the features of selected components for blocking elements. IV. TEST PROCEDURE AND EXPERIMENTAL RESULTS A. Test procedure The aim of the conducted tests is to quantify how voltage generation is affected when both generators are active. To this end, test procedure involves three steps: 1. both generators are connected, but only the 50 Hz generator is active; 2. both generators are connected, but only the highfrequency generator is active; 3. both generators are connected and active. In the following, the first two tests are indicated using the subscript “S”, whereas the third one is addressed by the subscript “FHF”. The amplitude of the fundamental component is chosen equal to 12 kV. The frequency of the high-frequency tone varies from 10 kHz to 150 kHz, while its amplitude is chosen as equal to 1 % of fundamental tone. The ratio error @ has been calculated comparing the amplitudes generated during the single-tone generation to the ones generated in the FHF test. Mathematically, the ratio error @ is defined as in (4): @ABCBDAE AE (4) where: • F>G> is the measured voltage at frequency  under the FHF test; • FH is the measured voltage at frequency  under the S test. As the only difference between the single-tone generation S test and the FHF test is the presence of both generators, the index @ estimates their mutual influence. B. Experimental results In the following, experimental results are provided and discussed. Fig. 3 provides the values of the errors @ associated to the generated high-frequency tones. With regards to Fig. 3, it can be observed that the mutual influence of the two generators is relatively low up to 50 kHz, because the error @ is less than 2 %. Up to 100 kHz, the error is below 3 %, whereas at 150 kHz it increases to 6 %. As the error curve has a negative polarity for all the frequencies, it means that, during FHF tests, HF tone amplitudes decrease due to the simultaneous presence of the other generator, with respect to S test. With respect to the error at 50 Hz (Fig. 4), it can be observed that it is much lower, and it is lower than ± 0.1 %. Moreover, the frequency of the high-frequency tone does not significantly influence the error at 50 Hz. Therefore, it can be concluded that the biggest influence is from the 50 Hz generator to the high-frequency one, whereas TABLE I. BLOCKING ELEMENTS Blocking element n. 1 Blocking element n. 2 C 2.5 nF 2 nF R 100 kΩ 50 Ω fcutoff ≅ 636 Hz ≅ 1.5 MHz Fig. 3. Impact of the simultaneous presence of both generators on the generated high-frequency tone vs. frequency Fig. 4. Impact of the simultaneous presence of both generators on the generated 50 Hz tone vs. frequency the 50 Hz generation is little affected by the mutual presence of both generators. C. Blocking components This subsection provides some results to evaluate the effect of the blocking elements. The residual high-frequency voltage measured by D1 has been compared to the full voltage measured by UD. Mathematically, it was calculated as in (2): JKLG> AM  ANM (2) Fig. 5 provides the values of JKLG> vs. the frequency of the generated high-frequency tone. It can be observed that the residual JKLG> decreases with frequency, because the higher the frequency, the farther you are from cutoff frequency. On the other side, the residual 50 Hz voltage JKL GO measured by D2 has been compared to the full voltage measured by UD. Mathematically, it was calculated as in (3): JKL GO AM# GO ANM GO (2) Fig. 6 provides the values of JKL GO vs. the frequency of the generated high-frequency tone. The cutoff frequency of the blocking element n. 2 is high enough to avoid significant 50 Hz voltage to be present at the input of the highfrequency amplifier. It can be observed that the residual 50 Hz voltage which falls across the high-frequency generator output is less than 0.002 % of the generated 50 Hz voltage. That ensures safety for the high-frequency amplifier and avoids that the 50 Hz current coming from the 50 Hz side and flowing into high-frequency amplifier is too high. V. CONCLUSIONS This paper presents a preliminary study about the implementation of a generation and measurement setup for the characterization of Voltage Transformers, for Medium Voltage grids, in the frequency range from power frequency (50/60 Hz) up to 150 kHz. The proposed generation system is based on the separate generation of the 50 Hz fundamental tone and the high-frequency components. The two generators are grounded and parallel-connected to obtain the composed voltage waveform. Two blocking elements have been added to the setup to prevent 50 Hz current from flowing into the high-frequency generator, and vice versa. Preliminary experimental results related to the mutual influence of the two generators are also discussed. It can be concluded that the biggest influence is from the 50 Hz generator to the highfrequency one, whereas the 50 Hz generation is little affected by the mutual presence of both generators. ACKNOWLEDGMENT The work here presented has received funding from EPM 22NRM06 ADMIT. This project 22NRM06 ADMIT has received funding from the EPM program co-financed by the Participating States and from the European Union’s Horizon 2020 research and innovation program. REFERENCES [1] “IEC 62488 Standard family, “Power line communication systems for power utility applications”. [2] P. Kotsampopoulos et al., "EMC Issues in the Interaction Between Smart Meters and Power-Electronic Interfaces," in IEEE Transactions on Power Delivery, vol. 32, no. 2, pp. 822-831, April 2017 [3] S. K. Rönnberg, A. G. -d. Castro, M. H. J. Bollen, A. Moreno-Munoz and E. Romero-Cadaval, "Supraharmonics from power electronics converters," 2015 9th International Conference on Compatibility and Power Electronics (CPE), Costa da Caparica, Portugal, 2015, pp. 539544. [4] Sarah K. Rönnberg, Math H.J. Bollen, Hortensia Amaris, Gary W. Chang, Irene Y.H. Gu, Łukasz H. Kocewiak, Jan Meyer, Magnus Olofsson, Paulo F. Ribeiro, Jan Desmet, “On waveform distortion in the frequency range of 2kHz–150kHz—Review and research challenges”, Electric Power Systems Research, Volume 150, 2017, Pages 1-10 [5] Y. Xia, J. Roy and R. Ayyanar, "Optimal Variable Switching Frequency Scheme to Reduce Loss of Single-Phase Grid-Connected Inverter With Unipolar and Bipolar PWM," in IEEE Journal of Emerging and Selected Topics in Power Electronics, vol. 9, no. 1, pp. 1013-1026, Feb. 2021. [6] “Z. Lu et al., "Medium Voltage Soft-Switching DC/DC Converter With Series-Connected SiC MOSFETs," in IEEE Transactions on Power Electronics, vol. 36, no. 2, pp. 1451-1462, Feb. 2021”. [7] E. O. A. Larsson, M. H. J. Bollen, M. G. Wahlberg, C. M. Lundmark and S. K. Rönnberg, "Measurements of High-Frequency (2–150 kHz) Distortion in Low-Voltage Networks," in IEEE Transactions on Power Delivery, vol. 25, no. 3, pp. 1749-1757, July 2010 [8] V. Khokhlov, J. Meyer, A. Grevener, T. Busatto and S. Rönnberg, "Comparison of Measurement Methods for the Frequency Range 2– 150 kHz (Supraharmonics) Based on the Present Standards Framework," in IEEE Access, vol. 8, pp. 77618-77630, 2020 Fig. 5. Percentage of high-frequency voltage measured by D1 compared to full high-frequency voltage measured by UD vs. frequency Fig. 6. Percentage of 50 Hz voltage measured by D2 compared to full 50 Hz voltage measured by UD vs. frequency [9] S. T. Y. Alfalahi et al., "Supraharmonics in Power Grid: Identification, Standards, and Measurement Techniques," in IEEE Access, vol. 9, pp. 103677-103690, 2021 [10] A. J. Collin, A. Delle Femine, C. Landi, R. Langella, M. Luiso and A. Testa, "The Role of Supply Conditions on the Measurement of HighFrequency Emissions," in IEEE Transactions on Instrumentation and Measurement, vol. 69, no. 9, pp. 6667-6676, Sept. 2020 [11] IEC 61869-6:2016, “Instrument transformers - Part 6: Additional general requirements for low-power instrument transformers”. [12] IEC 61869-14:2018, “Instrument transformers - Part 14: Additional requirements for current transformers for DC applications”. [13] IEC 61869-15:2018, “Instrument transformers - Part 15: Additional requirements for voltage transformers for DC applications”. [14] IEC 61869-1:2024, “Instrument transformers - Part 1: General requirements”. [15] M. Agazar et al., “Power Grids and Instrument Transformers up to 150 kHz: A Review of Literature and Standards”, Sensors, vol. 24, no. 13, p. 4148, Jun. 2024, doi: 10.3390/s24134148. [16] EPM 22NRM06 ADMIT, https://www.euramet.org/news-eventsseg/detailsnews?tx_news_pi1%5Baction%5D=detail&tx_news_pi1%5Bcontroll er%5D=News&tx_news_pi1%5Bnews%5D=1743&cHash=3dee4ab7 4ccbad5c668e69d243abf8fc [online on 5 June 2023] [17] G. D’Avanzo et al., "Theory and Experimental Validation of Two Techniques for Compensating VT Nonlinearities," in IEEE Transactions on Instrumentation and Measurement, vol. 71, pp. 1-12, 2022, Art no. 9001312 [18] G. Crotti, G. D’Avanzo, P. S. Letizia and M. Luiso, "The Use of Voltage Transformers for the Measurement of Power System Subharmonics in Compliance With International Standards," in IEEE Transactions on Instrumentation and Measurement, vol. 71, pp. 1-12, 2022, Art no. 9005912. [19] E. Mohns, J. Chunyang, H. Badura and P. Raether, "A Fundamental Step-Up Method for Standard Voltage Transformers Based on an Active Capacitive High-Voltage Divider," in IEEE Transactions on Instrumentation and Measurement, vol. 68, no. 6, pp. 2121-2128, June 2019. [20] M. Faifer, A. Ferrero, C. Laurano, R. Ottoboni, S. Toscani and M. Zanoni, "An Innovative Approach to Express Uncertainty Introduced by Voltage Transformers," in IEEE Transactions on Instrumentation and Measurement, vol. 69, no. 9, pp. 6696-6703, Sept. 2020. [21] A. Mingotti, L. Peretto and R. Tinarelli, "Effects of Multiple Influence Quantities on Rogowski-Coil-Type Current Transformers," in IEEE Transactions on Instrumentation and Measurement, vol. 69, no. 7, pp. 4827-4834, July 2020. [22] R. Stiegler and J. Meyer, "Impact of external influences on the frequency dependent transfer ratio of resin cast MV voltage instrument transformers," 2022 20th International Conference on Harmonics & Quality of Power (ICHQP), Naples, Italy, 2022, pp. 1-6 [23] J. Hällström et al., "Performance of a Modular Wideband HVDC Reference Divider for Voltages up to 1000 kV," in IEEE Transactions on Instrumentation and Measurement, vol. 64, no. 6, pp. 1390-1397, June 2015. [24] G. Crotti et al., "Characterization of Voltage Transformers for MV Applications Up to 150 kHz - A Preliminary Study," 2023 IEEE 13th International Workshop on Applied Measurements for Power Systems (AMPS), Bern, Switzerland, 2023, pp. 01-05, doi: 10.1109/AMPS59207.2023.10297174. [25] G. Crotti et al., "A Novel Generation and Measurement Setup for the Characterization of MV Voltage Transformers from 9 kHz up to 150 kHz," in IEEE Transactions on Instrumentation and Measurement, vol. 73, pp. 1-11, 2024, Art no. 9004711, doi: 10.1109/TIM.2024.3427768.