| 1. |
- de Oliveira, Roger Alves
(författare)
-
Applications of Unsupervised Deep Learning for Analysing Time-Varying Power Quality Big Data
- 2023
-
Doktorsavhandling (övrigt vetenskapligt/konstnärligt)abstract
- Continuous power quality monitoring allows grid stakeholders to obtain information about the performance of the network and costumer facilities. Moreover, the analysis of continuous monitoring allows researchers to obtain knowledge on power quality phenomena. Power quality measurements result in a large amount of data. Power quality data can be classified as big data, not only for its volume, but also for the other complexities: velocity, variety, and veracity. Manual analysis of power quality data is possible but time-consuming. Moreover, data reports based on standardized indexes and classical statistical techniques might hide important information of the time-varying behaviour in power quality measurements. Artificial intelligence plays a role in providing automatic tools for proper analytics of big data. A subset of artificial intelligence called machine learning has enabled computers to learn without explicit programming. Driven by the huge improvements in computer processing, a subset of machine learning based on multiple layers of artificial neural networks has been developed to tackle increasingly complex problems. The so-called deep learning applications teach themselves to perform a specific task by automatically extracting essential features from the raw data. Despite the possibilities of automatic feature extraction, most applications of deep learning to power quality are still the same as expert systems or earliest machine learning tools. Moreover, most of the applications are based on synthetic generated data and supervised techniques. In this context, the main motivation of this thesis is providing a new tool based on unsupervised deep learning to handle analytics of time-varying power quality big data. The unsupervised deep learning method proposed in this thesis combines a deep autoencoder with clustering for extracting patterns and anomalies in power quality big data. The deep autoencoder maps the original data to a compressed format that contains the principal features of the data. Automatic results are provided by the deep learning, and inferences can be obtained without requiring prior knowledge of deep learning. The outputs from unsupervised deep learning can serve as a guide for further data analysis, highlighting important time steps within large power quality datasets. By following these indications from the deep learning results, experts gain valuable insights into power quality phenomena, which can be referred to as "learning from deep learning". The interpretation of the deep learning results in this thesis allowed to making proper inferences for patterns and anomalies. For power quality measurements synchronised with 24-h, the results allowed making inferences concerning daily variations, seasonality, and the origins of power quality disturbances. For power quality measurements non-synchronised with 24-h, the results could be interpreted visually through the distribution of the patterns in a physical variable, such as the dynamic operating conditions of an electrical railway power system.An important contribution of this thesis concerns the physical interpretation of the phenomena is related to the anomalies in harmonics caused by geomagnetically induced currents. An interesting finding by applying the deep anomaly detection to measurements in the Swedish transmission grid is the damping of the anomalies caused by geomagnetically induced currents in the winter due to the heating load. This thesis also demonstrated that the signatures for anomalies in harmonic measurements in a Swedish transmission location are similar to the ones found in a low-latitude transmission location at the South Atlantic Anomaly due to geomagnetically induced currents. Moreover, by cross-checking the anomalies at the South Atlantic Anomaly with protection trips with undetermined causes, this thesis demonstrated that anomaly harmonics due to geomagnetically induced currents can cause protection mal trips.This thesis demonstrates that unsupervised deep learning can serve as an additional tool for compressing time-varying power quality big data into a more interpretable form. Despite the application of an unsupervised method, power quality experts remain significant in power quality studies. The main conclusion of this thesis is that unsupervised deep learning enhances the understanding of power quality experts and provides a complementary approach for analysing and extracting insights from time-varying power quality big data.
|
|
| 2. |
- Espin Delgado, Angela
(författare)
-
Propagation of Supraharmonics in Low-Voltage Networks
- 2022
-
Doktorsavhandling (övrigt vetenskapligt/konstnärligt)abstract
- The need for measures towards a sustainable use of energy has incited the proliferation of devices and systems for the efficient use of electricity. Energy-efficient appliances, equipment for the electrification of transportation, electricity generators from renewable energy sources, and communication protocols, e.g., for smart metering are sources of supraharmonic distortion in electrical networks. Supraharmonics are voltage and current waveform distortion in the frequency range from 2 up to 150 kHz.The increase in sources of supraharmonics in the last decades and the propagation of this type of distortion have triggered a variety of unwanted consequences (interference) in the electrical networks. Interference associated to supraharmonics such as audible noise, degradation or failure in the operation of electrical equipment, and breakdown of insulation materials, have been reported around the world. A standardized framework for supraharmonics as a power quality phenomenon that involves both grid operators and equipment manufacturers is needed to limit these interferences. The limits to be set shall not hinder the modernization of the electrical system and the consequential energy transition.There are gaps in the standardization framework for supraharmonics as a power quality phenomenon. The study of supraharmonics as a power quality parameter should consider variables that affect emission levels and propagation of supraharmonics. At the same time, an assessment of the severity of given supraharmonics levels regarding their consequences is needed to settle realistic reference levels. Deterministic methods have been generally used to study supraharmonic propagation but they might not be suitable when considering many possible scenarios.This research introduces forefront methods and results on the study of supraharmonics emission, propagation, and consequences. The study has two focal points: 1) to study the impact of the impedance of the grid and low-voltage devices on the emission and propagation of supraharmonics; 2) to assess the severity of propagated supraharmonics in terms of the characteristics of the distortion and the probability of interference. Experimental and theoretical case studies are built to carry out the research. Measured and synthetic signals representative of supraharmonic distortion present in low-voltage networks are used.The main results of this research are summarized as:The levels of emitted and propagated supraharmonics depend on the impedance of the grid, the emitting device and the neighboring devices. Resonance can lead to significant levels of supraharmonics anywhere in the grid. The variability and diversity of low-voltage devices lead to high uncertainty in the estimation of their impedance. Stochastic methods are recommended to assess the probability of interference.Different attributes of supraharmonics are responsible for different interference phenomena. Indications of the severity of supraharmonics attributes are given for three phenomena: audible noise, negative impacton residual current devices, and light flicker of LED lamps.This research contributes to the establishment of supraharmonics as a power quality phenomenon with standardized solutions. It introduces methods for the assessment of: 1) supraharmonic emission from installations needed to recommend planning levels; 2) supraharmonic propagation in low-voltage networks, and 3) the probability of interference needed to define reference levels.
|
|
| 3. |
- Gutierrez Ballesteros, Elena, 1996-
(författare)
-
On voltage fluctuations in low voltage grids – characterization, propagation and impact on LED lamps
- 2023
-
Doktorsavhandling (övrigt vetenskapligt/konstnärligt)abstract
- The work presented focuses on voltage fluctuations produced by household devices and their impact on the light output of LED lamps. Voltage fluctuations can produce unsteadiness on the light intensity of the lamps, so-called light flicker. An observer may or may not experience light flicker, depending on e.g., the severity of the voltage fluctuation, the lighting technology affected and the characteristics of the observer itself. The presence of voltage fluctuations in the grid is limited in the standards. One of the reasons for this limit is the impact the voltage fluctuations have on an incandescent lamp and by the irritability experienced by an average observer experiencing the resulting light flicker. New lighting technologies have been developed replacing the incandescent lamp e.g., halogen lamps, fluorescent lamps or most recently LED lamps. LED lamps are already in the market and are expected to dominate it in the future years.The impact that voltage fluctuations produce on LED lamps is studied in this thesis. To address this study, the factors other than voltage fluctuations that could affect the light output of LED lamps are studied, e.g., stabilization of the LED lamp, topology, impact of aging on the LED lamp or ambient disturbances.The study is focused on the voltage fluctuations produced by household devices. Measurements of household devices are performed to characterize the voltage fluctuations that they produce by e.g., their magnitude, rate of change, frequency or perception level by an average observer of light flicker on LED lamps.Different LED lamps have different responses on the light output subjected to a similar voltage fluctuation. This is a difficulty when considering LED lamps as a reference to measure their ritability experienced by an observer due to voltage fluctuations. The study offers a solution based on the characterization of the response of a generalized LED lamp.Uncertainties, such as the use of household devices or the occurrence of voltage fluctuations, are addressed using a stochastic approach, introducing the concept of risk to light flicker. Therefore, stochastic assessments of the risk and severity of light flicker are performed in a low voltage grid at different points of evaluation considering the propagation of voltage fluctuations between customers. In addition, the results of the severity of light flicker considering the response of a generalized LED lamp are compared to the severity obtained with an incandescent lamp.
|
|
| 4. |
- Hajeforosh, Seyede Fatemeh, 1988-
(författare)
-
Multiple Aspects of Dynamic Rating in the Power System
- 2022
-
Doktorsavhandling (övrigt vetenskapligt/konstnärligt)abstract
- Increased consumption and increased use of renewable energy make overhead lines and transformers more often congested. Dynamic rating (DR) uses a time-depending maximum permissible current instead of a long-term-fixed rating; it is an effective solution to upgrade the capacity of existing grid assets to minimize this congestion. Dynamic line rating (DLR) considers how the weather affects the thermal behavior of the conductor and therewith its rating; environmental parameters as well as conductor characteristics have to be considered. Likewise, dynamic rating of transformers (DTR) relies on a thermal assessment of the hottest spot in the transformer windings. The hot-spot temperature is dependent on the ambient temperature, loading, and transformer’s cooling system.In this thesis, a number of lesser-studied aspects of DR are studied: stochastic modeling of the rating; relations between protection, reliability and rating; risk assessment for line selection when using DLR; and increased transformer hosting capacity (HC) in the presence of solar photovoltaic (PV).In the first part of the work, the focus is on the stochastic modeling of DLR and its relation to the protection operation. Actual line rating is considered a random variable in the protection system for a more flexible decision-making. Different sources of uncertainties are modeled using suitable probability density functions. The method allows for a transparent trade-off between the risk of failure to take measures and the risk of unnecessary measures against overload. The results depicts that deterministic DLR could result in high probabilities of overloading or would require large safety margins. While, a stochastic approach will allow for both small margins and appropriate risks. A generic model is introduced to consider DLR reliability from two different viewpoints: errors and failures of components that would affect the calculation of the rating; and the impact of a DLR failure on the power system. The qualitative reliability study highlights that it is important to update the protection settings on a real-time basis. In the second part of the work, a risk assessment framework is proposed to select the minimum set of overhead lines for DLR implementation. Results show a possible increase of permissible hosting capacity (HC) for electric-vehicle charging by up to 80% (depending on the test system and initial data) with low interruption costs and reduced risk of congestion. Furthermore, the improvement in the transformer HC by using DTR has been quantified. The results indicate that depending on the HC performance indices, the transformer can be loaded beyond the normal operational limits up to 35% to host PV and up to 100% for increased consumption.
|
|
| 5. |
- Nakhodchi, Naser, 1981-
(författare)
-
On Harmonics in Low-voltage Networks
- 2023
-
Doktorsavhandling (övrigt vetenskapligt/konstnärligt)abstract
- On the road to reducing global warming, the use of renewable energy sources and efficient use of electricity are among the key aspects. The increasing number of energy efficient appliances such as LED lamps, booming demand for electric vehicles (EVs), and growing penetration of distributed energy resources such as photovoltaic (PV) systems in low-voltage (LV) networks are expected to affect the power quality in the entire electric power system and specifically in the LV network, where such new devices are connected. Waveform distortion, mainly expressed by the harmonic components, is one of the topics within power quality that is highly affected by the introduction of such new devices. However, there is a lack of publications discussing the existing level of voltage harmonics in LV networks or addressing the origin and transfer of harmonics in LV and MV distribution networks. This highlights the need for more research in this field.To evaluate the ability of the network to host new sources of harmonics, the existing harmonic voltage and current levels as well as the impact of these new sources on those levels should be investigated. Harmonic levels are determined by emission from harmonic sources, the propagation from other harmonic sources, and the aggregation between the contributions from different sources. Studies on harmonic emission from a variety of different individual devices under different conditions have already been carried out. However, limited knowledge is available about the harmonic aggregation and propagation in LV networks. This study aims to improve the understanding about the behaviour of harmonics in LV networks covering both aggregation and propagation.In the first part of this work, the impact of the MV network and remote LV loads on the harmonic voltage in the LV network are examined. Simulation results have revealed that for frequencies below the resonant frequency of the local LV network the harmonic voltage levels mainly are determined by aggregated emission of the whole distribution network (both LV and MV) rather than by the emission from local LV loads. Furthermore, a graphical method is introduced for harmonic propagation studies, using measurements but without the need for accurate synchronized measurements.In the second part of this work, the aggregated emission from a group of EV fast chargers is examined. A stochastic method, based on Bayesian statistics and harmonic correlation, was used to include uncertainties in harmonic hosting capacity calculation for an EV charging station equipped with fast chargers. The impact of MV network and remote LV loads on harmonic hosting capacity is investigated. It is also shown that harmonic hosting capacity studies are needed; and details of the distribution network must be included to get an accurate estimation of the harmonic hosting capacity.Finally, an alternative method for time aggregation of harmonic phase angle is proposed in this work.In general, this work contributes to reducing the research gaps recognized in harmonic analysis in the LV networks considering propagation and aggregation utilizing both simulation and measurement.
|
|
| 6. |
- Nömm, Jakob
(författare)
-
Power quality analysis and techno-economic modeling for microgrids
- 2021
-
Doktorsavhandling (övrigt vetenskapligt/konstnärligt)abstract
- The work done in this thesis considers microgrids from two different aspects. Power quality and techno-economics of microgrids. Detailed power quality measurements have been made at a single house hydrogen-solar microgrid that consists of state-of-the-art energy efficiency technology, energy production and energy storage. The microgrid can both connect to the grid and operate in islanded operation. The power quality is quantified from these measurements where several power quality parameters during islanded operation go beyond the limits set by standards such as EN 50160 and IEEE 519-2014. The effect on connected equipment from both frequency variations and voltage quality is also discussed. Four new performance indexes are presented in the thesis that are based on apparent impedances. The first with the name PHIPI quantifies how much the harmonic voltage magnitude changes with an increase in harmonic current magnitude on the same phase. The second with the name SHIPI quantifies how much the harmonic voltage magnitude changes with an increase in harmonic current magnitude on another phase. The third with the name AHSI uses the harmonic voltage and current magnitudes of all phases to create a single performance parameter expressed as an apparent impedance for the system. The fourth with the name ARMSSI quantifies the phase RMS voltage drop for a certain phase RMS current rise in terms of an apparent impedance. The thesis also shows techno-economic modeling with times series energy flow to study the investment risks related to consumption changes in a standalone microgrid. The results show that consumption changes are an important parameter when designing a standalone microgrid and that the risk can be mitigated with changes to the system design, but at a larger system cost. The projected cost reduction until the year 2050 for standalone hydrogen based microgrids and some risk aspects with hydrogen based microgrids are also discussed in the thesis.
|
|
| 7. |
- Rönnberg, Sarah
(författare)
-
Emission and interaction from domestic installations in the low voltage electricity network, up to 150 kHz
- 2013
-
Doktorsavhandling (övrigt vetenskapligt/konstnärligt)abstract
- This thesis work has focused on conducted emission (up to 150 kHz) from common lowvoltage appliances. The emphasis has been on equipment that contributes to a sustainable energy system: photovoltaic (PV) installations and energy-saving lamps (LED lamps). The frequency components present in the grid in addition to the fundamental 50 Hz component can be divided into harmonics (up to 2 kHz in a 50 Hz system) and supraharmonics (2 kHz to 150 kHz). These frequency components are partly the effect of normal operation of equipment due to power-electronic converters and the switching technique used. Power line communication, PLC, is an important source of frequency components in the range 9 to 95 kHz. Even though from an equipment viewpoint there is no difference between a signal used for communication and a signal that is a residue from a switching circuit, PLC is a useful signal for operation of the grid and for communication with electricity meters. The amplitude of the communication signal is in in almost all cases higher than the emission from any other equipment connected to the grid.Understanding the different types of interaction between PLC and end-user equipment has been a major part of this work. Five types of interaction have been identified; some negative for PLC, some negative for end-user equipment. An important conclusion from this part of the work is that loss of communication with PLC, as is often reported with remote reading of electricity meters, is not due to emission by end-user equipment but due to the EMC filter of the end-user equipment providing a low-impedance path. The understandings acquired from the work with PLC have been applied to other types of emission as well. Supraharmonics from individual devices, above about 10 kHz, flow mainly to neighboring devices, not into the grid. This behavior was found by laboratory experiments and confirmed by other studies as well. A circuit-theory model has been developed that explains this behavior. The EMC filters are shown to be the main cause of this behavior. Other configurations of those filters may result in a larger flow of emission towards the grid.One type of appliance that has been introduced recently is the LED lamp. LED lamps come in different designs with different emission spectra. A possible distinction is between lamps with high levels of low-order harmonics (up to a few 100 Hz) and those with high levels of supraharmonics. Restricting the emission in the lower frequency range, through standardization, could result in higher distortion levels at higher frequencies. Replacement of incandescent lamps by CLF and LED lamps is not expected to result in a noticeable increase in harmonic voltage and current levels in the grid. This has been shown through several laboratory experiments and field experiments.Emission from PV inverters is low at low-order harmonics, this have been shown by a number of measurements both on single phase connected installations, so called rooftop installations and larger, three phase connected installations (20 kW). In addition to emission in the low frequency range, PV inverters emit at their switching frequency, e.g. around 16 kHz. The emission at this frequency is shown to vary by a factor of 5 or more, depending on the presence of neighboring equipment. This was shown by measurements and has been explained by a circuit-theory model. This thesis work has resulted in further understanding on the emission from PV panels and energy-saving lamps and on the propagation of conducted emission from common household appliances. This work is an important contribution to the research on distortion of voltage and current in the frequency range 2 to 150 kHz.
|
|
| 8. |
- Sakar, Selcuk
(författare)
-
Modelling and Interference Analysis of AC-DC Converters for Immunity to Voltage Disturbances up to 150 kHz
- 2021
-
Doktorsavhandling (övrigt vetenskapligt/konstnärligt)abstract
- The function of electrical and electronic equipment is challenged by the high frequency emission originated from the more use of switching-based power electronic equipment and Power line communication (PLC) signaling. Investigations have shown that interferences occur in the frequency range between 2 kHz and 150 kHz, e.g., error in smart meter reading, audible noises, performance degradation and even malfunction of the equipment. This study aims to model and analyze the interference mechanism and gain the knowledge of equipment behavior/immunity under voltage disturbance from 2 kHz up to 150 kHz, so called “Supraharmonics.”Experimental studies were conducted on different type of AC-DC converters, LED drivers, computer power supplies, Active Power Factor Correction (APFC) circuits (pre-regulator), single stage and double stage converters such as flyback and buck topologies. AC-DC converters were exposed to Supraharmonics (SHs) according to IEC 61000-4-19. Interferences, that occur in the input voltage/current and dc-link, were observed and analyzed through a mathematical model of converter circuit. When the AC-DC converter is exposed to SHs, three phenomena occur that 1) the conduction and blocking time of the diodes, in the bridge rectifier circuit, are deviated, 2) intermittent conduction takes places in the beginning and of diodes conduction, 3) reverse-recovery of the diodes happen that impact the dc-link voltage of the converter. Those behaviors are verified in a functional model and validated through the experiments. It is also shown that dc-link voltage metrics, peak-to-peak and average value, are impacted when SHs are present at the terminal of the converter. Further, it is revealed that if the SHs are nonsynchronized to the power system frequency, the converter due to uneven deformation of the diode conduction /blocking time generates interharmonic currents. The generated interharmonic frequency can be precisely determined by using frequency modulation equation. Result shows that generated interharmonic magnitudes are negligible if the applied SH frequency is above 16 kHz. In order to quantify the impact of SHs on dc-link capacitor, an immunity model that links the intermittent conduction and the dc-link capacitor current is established, analytically. By this means, it is ensured to avoid the degradation of the dc-link capacitor’s lifetime by defining the SHs voltage limits.The analysis and modelling studies proves that SHs at the AC-DC converter terminal can cause deviations in the function of equipment both in short term and long term., e.g., increased light intensity in LED drivers, reduced peak-to-peak voltage in dc-link, interharmonic injection and reduced estimated lifetime of the dc-link capacitor. Those impacts originates from the reaction of the diodes and non-linear circuit configuration. Functional models presented in this study are able to express the reasoning and occurrence of the impacts. Established immunity model analytically gives guidance on how immune converters can be achieved in the design stage of power electronic circuits. This study provides deep insight on how AC-DC converters behaves under SH emission. Standard committees and converter manufacturer, to achieve immune and reliable power system and equipment for the future’s systems, can use this knowledge.
|
|
| 9. |
- Sutaria, Jil
(författare)
-
On the supraharmonics in single-phase and three-phase installations
- 2023
-
Doktorsavhandling (övrigt vetenskapligt/konstnärligt)abstract
- Society has increasingly started depending on the continuity of internet services, the interruption can be crucial for vital societal functions. It is therefore important to maintain continuous and reliable operations of e.g., installations housing sensitive IT (Information Technology) loads. The establishment of a data center attracts huge investments, thereby boosting the economy. The electrical load within the data center is more structured with a similar type of load, i.e., power supplies supplying the servers connected repetitively. They also use power-conditioning equipment such as an uninterruptible power supply (UPS), which ensures uninterrupted power to the servers. The power supplies use switching techniques to modulate and condition the power supplied to the IT load. The switching residues from the power supplies are often in the frequency range from 2 kHz to 150 kHz, known as supraharmonics. The switching frequency of power electronic converters and power line communication are identified as the main contributors to supraharmonics. The increase in the number of power electronic converters in the power system has sparked interest in supraharmonics. Limited standards are available on the magnitude of supraharmonics allowed in the system. For installations that house multiple power converters, prior knowledge about the propagation and consequences of the supraharmonics can be an advantage in increasing their reliability. This work is focused on supraharmonics within data centers, but the results of the study can be applied to other installations.The data centers are usually three-phase installations where single-phase devices, i.e., servers are distributed over three phases. To understand the holistic view of the system, it is important to understand the spread of supraharmonics in phase, neutral, and protective earth conductors. A perspective on the behaviour of supraharmonics in three-phase installation is presented in this work. The focal point of the work is the summation of supraharmonics in the neutral conductor and conductor crosstalk. The behaviour of the device connected in single-phase installation and when the same devices are spread over three phases may differ depending on the design of the installation. Since the focus is on a greater number of devices, analysis for the change in the magnitude of supraharmonics in a neutral conductor with the change in the number of devices is also undertaken. The IT workload in a data center changes continuously, thus a constant unbalance in the electrical load between the three phases can exist. The thesis aims to understand the factors influencing the change in supraharmonic magnitude and propagation in phase and neutral conductors according to a balanced and unbalanced voltage supply and loading conditions. The presence of a supraharmonic emitting load in one phase can cause interference on the adjacent phase due to conductor crosstalk. Crosstalk leads to the propagation of supraharmonics between the phases. The work focuses on the parameters influencing the crosstalk phenomena for supraharmonics. Among the parameters studied, the type of cable used in an installation and the cable length is dominating. The measurement of supraharmonic emission from the device connected as single-phase and connected as three-phase would differ because of crosstalk.A large number of servers in data centers connected close by may lead to an increased leakage current. The composite leakage current can comprise subharmonics, harmonics, and supraharmonics. This can cause unwanted tripping of the residual current device (RCD), which is installed for human protection. The work aims to understand the impact of the composite residual current on the operation of the residual current devices.The main contributions of this work are summarized as:• A sensitivity analysis is presented based on the parameters impacting the propagation of supraharmonics for different operating modes of a UPS.• A characterization of supraharmonics emission from different devices according to their time and frequency varying behaviour.• A mathematical model to predict the change in the supraharmonic emission in the neutral conductor.• An increased understanding of the behaviour of supraharmonics for constant powerloads under voltage reduction conditions in three-phase installations.• Introduction and definition of the terms induced primary emission and induced secondary emission to explain conductor crosstalk for supraharmonic emission.• A mathematical model to show the factors influencing the magnitude of induced primary emission and induced secondary emission.• The frequency response of RCDs of type AC, A, B, and F under residual currents comprising pure tones and composite currents is presented by measurements andmodeling.
|
|
