Power Quality in a 110 kv DSO System Measured by Means of Non- Conventional Instrument Transformers

Transcription

1 Presentation 06.1 Power Quality in a 110 kv DSO System Measured by Means of Non- Conventional Instrument Transformers Dr. Philipp Werdelmann / Christoph German, Westnetz, Germany Abstract This paper presents power quality measurements within a 110 kv substation of WESTNETZ GmbH, Germany. Measurements are performed by making use of an RC voltage divider and a broadband acquisition unit. The intention is to find out the characteristics of both, steady state variations as well as transient disturbances. The obtained measurement results are evaluated and the possible impact on grid operability and the influence on equipment stress are discussed subsequently. 1 Introduction and Motivation Providing a high power quality is becoming more important concerning all stakeholders of energy grids, i.e. consumers, generators and of course operators. The term itself is often characterized by the quality of voltage supply being influenced by different kinds of disturbances or distortion. The analysis of Power Quality is often focused on most in LV and MV grids but only partially in HV grids which is certainly due to the different system impedance and the smaller amount of non-sinusoidal infeeders or consumers which are directly connected to the HV grid. However, this reason may change as the development of modern energy grids provides more decentralized and even more volatile generation of power, e.g. from renewable sources, as well within 110 kv distribution systems. Other effects which are due to more dynamic phenomena like control reaction, fault incidents or operational actions deliver additional impacts on Power Quality as well. 1.1 Today Transition in Distribution Grids and Future Challenges The present transition of energy grids in Germany leads to more decentralized generation with more volatile feed-in. Thus, a bidirectional power flow is challenging the grid infrastructure, which was intentionally planned and designed for a top-down, unidirectional power flow decades ago (Figure 1:). Especially for the distribution grids there are a lot of challenges the operators in Germany have to deal with today as well as in the future. On the one hand, the short-circuit capacity is getting lower as large rotating generators leave market, which again leads to a decrease in robustness and in system flexibility. On the other hand, more dynamic reaction is needed to operate the grid in a stable mode. Figure 1: Today s Development of Energy Grids Leads to Decentralization and Bidirectional Power Flow 1.2 Impacts on Power Quality and Infrastructure Reliability The different kinds of power quality variations can be classified into the following categories [1]: Steady State Variations Transient Disturbances Steady Sate Variations describe harmonic distortion and normal RMS voltage variations and are often evaluated on a larger time basis within days or weeks. Disturbances cover transient phenomena like surge or burst impulses. Transient disturbances can be the consequence of switching events, electric faults or induction phenomena like lightning strikes. These kinds of disturbances are usually detected by triggering on specific criteria, e.g. exceeding a voltage or rate-of-change threshold. Short-time deviations from nominal or rated values do also fall into this category [2]. The intention for investigating Power Quality in this case was to find out about: The level of harmonic distortion on 110 kv Voltage stability and level of unbalance Frequency deviations Criticality of transient effects Possible stress for equipment Any correlation of these effects, if applicable Steady state variations, e.g. harmonic distortion, are able to stress critical equipment like transformers due to increased dissipation losses (eddy currents, saturation effects, etc ). Another impact regarding

2 Presentation 06.2 unbalance and THD can be an overload of the neutral phase. Besides this, transient disturbances bear the risk of spontaneous and unplanned outages due to sudden equipment damage. Effects of damage caused by repetitive transients may also be a reason for advanced ageing and reduction in lifetime. 1.3 Standards and Guidelines In general, Power Quality aspects are described in the international standard IEC Electromagnetic Compatibility (EMC). In this class, specifically IEC holds guidelines and information about power quality measurement methods [3]. For this work, measurement techniques have been chosen according to the mentioned standard. Concerning limits and maximum values for power quality assessment the standard IEC addresses this topic, however, this standard is limited to LV and MV only. The European Standard EN defines characteristics for the usually expected variations and gives reference values which help to assess the detected Power Quality deviations [4]. Furthermore, for HV systems there are Technical Rules for the Assessment of Network Disturbances, which define requirements for consumers and generators connected to the HV network [5]. 2 Measuring System and Setup In order to obtain measurement results which are representative for a specific state of grid operation the location for installing the measurement equipment should be of specific characteristic as well. This could be either a substation located nearby urban areas with high power consumption and distinctive load curves or a more rurally oriented substation location with high amount of volatile power feed-in. In this work, the latter option was taken, which is to be followed by the first one in near future. Measurements were taken during a period of three month beginning in July Substation Location For performing measurements in a rural but windy location the 110 kv substation Büren within the WESTNETZ operated area was chosen (see Figure 2:). In this area there are several wind parks connected directly to 110 kv and 30 kv lines. Figure 2: 110 kv Substation Büren is Operated by WESTNETZ (Blue Area) in the West of Germany 2.2 Equipment and Methodology For 110 kv measurements WESTNETZ mostly operates inductive instrument transformers. These transformers are designed for 50 Hz applications without the need for providing signals in a much wider frequency range. Therefore, for detecting higher harmonics as well as transients a voltage transformer design was needed which shows higher cut-off frequencies [6] [7]. A Pfiffner RC-divider was chosen providing highest accuracy with a maximum amplitude error of 0.2% up to 10 khz. A comparable design has already been used for PQ measurements in higher voltage levels [8]. These devices were mounted in a spare line field within substation Büren (see Figure 3:). The secondary output signals are picked up by a power quality analyzer ( PQBox200, a.eberle) with a sampling rate of 40 khz. The PQBox allows different options for signal data acquisition like oscilloscope, transient recorder and RMS monitor functions. By software parametrization the trigger and threshold values were adjusted during commissioning on site. Measurements are performed during a period of three months: The data acquisition unit continuously captures 10ms RMS values and is triggered by transient events to record the respective data. This storage is read out approximately every 2 weeks and post-processing of the raw data is done via MATLAB. Additionally, data from the grid control system was obtained and incorporated into evaluation, e.g. wind power measurements or grid operation protocols.

3 Presentation 06.3 (higher average value and less deviation) as well as by a lower voltage unbalance in total. It is more difficult to assess the THD values given in the mentioned interval, as not only feed-in of wind power might have a significant influence (due to harmonics caused by inverter units) but also the load characteristics of consumers. The THD value does not allow to separate this from one another clearly. However, the measured THD value is dominated by the 5 th and 7 th harmonic, which can be partially explained by inverter units of wind, photovoltaic or loads. The harmonic distortion has been measured up to the 50 th harmonic, which is presented in Figure 5:. Limit values as given in [4] are 5% for the 5 th and 4% for the 7 th order harmonic. Figure 3: RC Divider (Pfiffner ROF 123) Mounted in a 110 kv Substation for PQ Measurements 3 Measurement Results and Evaluation 3.1 Steady State Variations According to [4], a 7-day-intervall was evaluated and depicted in Figure 4: for RMS values of voltage, unbalance, frequency and THD. As the region around Büren is dominated by wind power generation, the average sum for the nearby wind park power feed-in is given isochronally. Figure 5: 24h Harmonic Distortion Measured in 110 kv Substation Büren on Out of this 7-day period, the (Saturday) is of specific interest, because the load is quite low on this weekend. Furthermore, there was much wind power available at this particular day. Figure 6: gives the 24-h data concerning the mentioned parameters for this day. Figure 4: 7d-Recordings of Voltage, Unbalance, Frequency, THD and Active Wind Power on Figure 4 illustrates, that all the measured values are far below the maximum allowable limits. The maximum average value of the voltage (117 kv) is found approximately 6% higher than the nominal voltage. This value was reached on a Sunday morning (very low electricity demand), as there was also heavy wind presence in that region. This effect is also reflected by the frequency Figure 6: 24h-Recordings of Voltage, Unbalance, Frequency, THD and Active Wind Power on (Sat.)

4 Presentation 06.4 On this day, wind power in total but also regarding the rate of change played a major role. Nevertheless, the Power Quality in the 110 kv distribution system was still in very good condition. On Wednesday, , (see Figure 7:) there was much more electricity demand pointed out by both, the voltage as well as the THD value in the beginning of working time that day (between 05:00 and 09:00 a.m.): voltage drops approximately by 3...4% until voltage regulation actions are activated (stepping of HV transformer OLTC) and total harmonic distortion consistently rises but is still far away from reaching limit values. As THD value decreases again during the afternoon hours, although wind power reaches its maximum in this time. This is shown with load deviations in this region, which have a stronger influence on Power Quality than with inverter operated generating units. qualitatively higher at the voltage zero crossing than at the voltage peak. Thus, if circuit breaker switching would have happened in the voltage maximum, the resulting overvoltage would have been smaller in this example event. Figure 8: Oscilloscope Recording of Circuit Breaker Switching on Another typical and spontaneous phenomena are earth faults. Figure 9: illustrates an oscilloscope recording of a self-extinguishing earth fault with a duration of approximately 20 ms which was initiated on phase L1. The transient consequence of this event was a relative voltage peak of ca. 50 kv on phase L3 as the star-point is shifted. Figure 7: 24h-Recordings of Voltage, Frequency, THD and Active Wind Power on (Wednesday) 3.2 Transient Disturbances During the measurement period there were some transient events triggered by the acquisition unit. Most often these were switching events followed by earth and line faults. Depending on attenuation effects due to distance, impedance and grid status the resulting overvoltage appears in different intensity. Sudden cut-off of inductive currents will additionally result in steep voltage impulses. As an example, Figure 8: shows the triggered voltage recording of a circuit breaker opening event detected in substation Büren. The relative voltage peak on phase L2 is about 45 kv and higher compared to L1 and L3. This voltage, of course, is depending on the inductive component of the current which is Figure 9: Oscilloscope Recording of a Transient Earth Fault Event on A very interesting incident event happened on , when a sustaining earth fault occurred on phase L3 (see Figure 10:). This fault developed to a line-to-line fault between L2 and L3 after 1 second, until 150 ms later a successful automatic reclose cleared the fault event.

5 Presentation 06.5 However, higher electric field stress will cause additional ageing to the devices and the dielectric system which is depending on different parameters, e.g. rate-of-change (du/dt) and duration. The presented work will be carried on in additional substations and regions covering different characteristics like load profiles, urban structures etc. to widen the basis of information about power quality in 110 kv distribution grids and possible impacts on equipment stress and grid planning strategies. Literature [1] Melhorn, Christopher; McGranaghan, Mark: Interpretation and Analysis of Power Quality Measurements; Elektrotek Concepts, Inc.; Knoxville, Tennesee Figure 10: Oscilloscope Recording of Sustaining Earth Fault Turning into a Line-to-Line Short Circuit before Being Successfully Reclosed Again, the earth fault caused a star-point shifting leading to a phase-to-neutral voltage rise of 3 on the healthy phases L1 and L2. The line-to-line fault was detected by the protection relay and a successful reclose could be performed in order to clear the fault. 4 Conclusions and Outlook In general, steady state as well as transient disturbances have the ability to cause damage to electrical equipment including, generators, operators and consumers. The performed measurements were able to show, that harmonic distortion is no crucial issue within the 110 kv system which was subject to investigation. There are trends to be seen in indirect correlation between electricity demand and derivations in frequency, unbalance and THD. Transients disturbances show significant voltage peaks which, however, is not unusual for normal grid operation, as fault events or switching are part of everyday grid operation. The question is, if the need for performing switching actions more often might increase in the future. HV equipment in substations has to be designed and tested for fault events, providing that protection relays will prevent serious damage due to overcurrent for example. Electric stress resulting from occurring higher voltages as shown in this work is surely within the limits of what type and routine tests do have to cover. [2] CIRED: Special Report Session 2: Power Quality and Electromagnetic Compatibility; 23 rd International Conference on Electricity Distribution; Lyon; June 2015 [3] DIN EN Elektromagnetische Verträglichkeit (EMV), Teil 4-30: Prüf- und Messverfahren Verfahren zur Messung der Spannungsqualität, Sept [4] DIN EN Merkmale der Spannung in öffentlichen Elektrizitätsversorgungsnetzen, Feb [5] Forum Netztechnik/Netzbetrieb im VDE (FNN): Technische Regeln zur Beurteilung von Netzrückwirkungen Ergänzungsdokument zur Beurteilung von Anlagen für den Anschluss an Hochspannungsverteilernetze.; 1st Ed [6] EURAMET: Publishable JRP Summary Report for ENG61 FutureGrid: Non-Conventional voltage and current sensors for future power grids; June 2015 [7] Meyer, J.; Stiegeler, R.; Klatt, M.; Elst, M.; Sperling, E.: Accuracy of Harmonic Voltage Measurements in the Frquency Range up to 5 khz using Conventional Instrument Transformers; 21 st International Conference on Electricity Distribution; Frankfurt; 6-9 June 2011 [8] Sperling, Erik; Schegner, Peter: A Possibility to Measure Power Quality with RC-Divider; 22 nd International Conference on Electricity Distribution; Stockholm; June 2013

6 Presentation 06.6 About the Authors Dr. Philipp Werdelmann received his electrical engineering diploma at Technical University Dortmund in After this he gained experience at the chair of High Voltage Engineering and EMC as research assistant until he finished his PhD in In that year he started working for RWE in the field of Power Plant Planning & Approval as project manager and technical expert for generators, transformers and electrical drives in Germany, UK, NL and Eastern Europe. In 2013 he joined WESTNETZ GmbH, RWE s distribution system operator in Germany. He now is responsible as team leader for HV equipment, product management, electrical and structural engineering and technical specification of substation design within the HV Substation Technology department. Christoph German received his electrical engineering diploma at Technische Fachhochschule Bochum in In that year he started working for RWE in Germany in the field of distribution and transmission grids as technical expert for disconnectors and surge arresters in 110 kv substations. He now is the technical product manager for current and voltage instrument transformers in RWE s distribution system operator Westnetz GmbH in Germany. The product management contains the whole lifecycle of the equipment including the technical specification, pre-qualification of suppliers, acceptance test, operation and disposal.