This Ph.D. thesis proposes an approach to classify faults that commonly occur in a High Voltage Direct Current (HVDC) power system. These faults are distributed throughout the entire HVDC system. The most recently published techniques for power system fault classification are the wavelet analysis, two-dimensional time-frequency representation for feature extraction and conventional artificial neural networks for fault type identification. The main limitation of these systems is that they are commonly designed to focus on a group of faults involved in a specific area of a power system. This thesis introduces a framework for fault classification that covers a wider range of faults. The proposed fault classification framework has been initiated and developed in the context of the HVDC power system at Manitoba Hydro, which uses what is known as the TranscanTM system to record and archive fault events in files. Each fault file includes the most active signals (there are 23 of them) in the power system. Testing the proposed framework for fault classification is based on fault files collected and classified manually over a period of two years. The fault classification framework presented in this thesis introduces the use of the rough membership function in the design of a neural fault classification system. A rough membership function makes it possible to distinguish similar feature values and measures the degree of overlap between a set of experimental values and a set of values representing a standard (e.g., set of values typically associated with a known fault). In addition to fault classification using rough neural networks, the proposed framework includes what is known as a linear mean and standard deviation classifier. The proposed framework also includes a classifier fusion technique as a means of increasing the fault classification accuracy.

This Ph.D. thesis proposes an approach to classify faults that commonly occur in a High Voltage Direct Current (HVDC) power system. These faults are distributed throughout the entire HVDC system. The most recently published techniques for power system fault classification are the wavelet analysis, two-dimensional time-frequency representation for feature extraction and conventional artificial neural networks for fault type identification. The main limitation of these systems is that they are commonly designed to focus on a group of faults involved in a specific area of a power system. This thesis introduces a framework for fault classification that covers a wider range of faults. The proposed fault classification framework has been initiated and developed in the context of the HVDC power system at Manitoba Hydro, which uses what is known as the TranscanTM system to record and archive fault events in files. Each fault file includes the most active signals (there are 23 of them) in the power system. Testing the proposed framework for fault classification is based on fault files collected and classified manually over a period of two years. The fault classification framework presented in this thesis introduces the use of the rough membership function in the design of a neural fault classification system. A rough membership function makes it possible to distinguish similar feature values and measures the degree of overlap between a set of experimental values and a set of values representing a standard (e.g., set of values typically associated with a known fault). In addition to fault classification using rough neural networks, the proposed framework includes what is known as a linear mean and standard deviation classifier. The proposed framework also includes a classifier fusion technique as a means of increasing the fault classification accuracy.

High Voltage Direct Current (HVDC) transmission systems continue to be an excellent asset in modern power systems, mainly for their ability to overcome the problems of AC transmission, such as the interconnection of asynchronous grids, stability of long transmission lines, and use of long cables for power transmission. In the past 20 years, Voltage Source Converter (VSC)-HVDC transmission systems were developed and installed in many projects, thereby adding more operational benefits to DC transmission option, such as high controllability, ability to supply weak networks, and reduced converter reactive power demand. Nevertheless, VSC-HVDC transmission suffers from the disadvantages of high losses and cost. In this research, a hybrid HVDC employing a Line Commutated Converter (LCC) as rectifier and a VSC as inverter is used to supply a passive network through a DC cable. The hybrid system is best suited for unidirectional power transmission scenarios, such as power transmission to islands and remote load centers, where the construction of new transmission lines is prohibitively expensive. Control modes for the rectifier and inverter are selected and implemented using Proportional Integral (PI) controllers. Special control schemes are developed for abnormal operating conditions such as starting at light load and recovering from AC network faults. The system performance under steady state and transient conditions is investigated by EMTP-RV simulations. The results show the feasibility of the hybrid system. / UOIT

With the great possibility of offshore wind power that can be installed in the world seas, offshore wind power is starting to get and important source of energy. The growing sizes of wind turbines and a growing distance to land, makes the choice of transmission alternative to a more important factor. The profitability of the transmission solution is affected by many parameters, like investment cost and power losses, but also by parameters like operation & maintenance and lead time of the system. The study is based on a planned wind farm with a rated power of 1 200 MW and at a distance of 125 km to the connection point. Four models have been made for the transmission network with the technology of HVAC, HVDC and a hybrid of both. The simulation program used is EeFarm II, which has an interface in Matlab and Simulink. The four solutions have been compared technically, with difficulties and advantages pointed out and also economically, with the help of LCOE, NPV and IRR. Costs, power losses and availability of the wind turbines and intra array network are not included in the study. The result of the simulations implies that the HVAC solution is the most profitable with the lowest Levelized Cost of Energy and highest Net Price Value and Internal Rate of Return. The values are 25.11 €/MWh, 387.60 M€ and 15.32 % respectively. A HVDC model with just one offshore converter station, has a LCOE close to the HVAC solution, but with a more noticeable difference in NPV and IRR (25.71 €/MWh, 300.76 M€ and 14.84 % respectively). A sensitivity analysis has been done, where seven different parameters have been changed for analysing their impact on the economic result. The largest impact made was by a change in investment cost and lead times. The results imply that with a structure of the transmission network as for the models, and with similar input data, the break point where a HVDC solution is more profitable than a HVAC solution is not yet passed at a distance of 125 km from the connection point. With an evolving technology in the field of HVDC, a shorter lead time and lower investment cost could mean that a HVDC solution would be more profitable at this distance. Difficulties for a HVAC solution with more cable required, like bigger land usage and cable manufacturing as a bottle neck, could make an important factor tough while making a decision. / Med den stora potentialen hos världens hav, börjar havsbaserad vindkraft bli en betydande energikälla. Den ökande storleken på vindkraftsturbinerna tillsammans med de ökade avstånden mellan vindkraftsparkerna och land, gör att transmissionslösningen blir en mer betydelsefull komponent. Flera olika parametrar kan vara avgörande för transmissionslösningens lönsamhet, som investeringskostnad och effektförluster, men också saker som drift & underhåll och projektets ledtid. Studien är baserad på en planerad vindkraftspark med en märkeffekt på 1 200 MW och på ett avstånd på 125 km till anslutningspunkten. Fyra modeller av transmissionssnätet har gjorts, där tekniken har bestått av HVAC, HVDC samt en blandning av dessa. Simuleringarna har gjort i EeFarm II, ett program baserat på Matlab och Simulink. De fyra modellerna har jämförts tekniskt, med för- och nackdelar poängterade, och även ekonomiskt med hjälp av LCOE, NPV och IRR. Kostnader, effektförluster och tillgängligheten för vindkraftsturbinerna och internnätet i vindkraftsparken är inte inkluderade i studien. Resultaten av simuleringarna visar på att HVAC-lösningen är den mest lönsamma, med lägst Levelized Cost of Energy och högst Net Price Value och Internal Rate of Return. Värdena för dessa är 25,11 €/MWh, 387,60 M€ respektive 15,32 %. En HVDC-lösning med enbart en DC-plattform och likriktarstation för hela märkeffekten, har en LCOE inte långt ifrån HVAC-lösningen, men med en lite större skillnad i NPV och IRR (25,71 €/MWh, 300,76 M€ respektive 14,84 %). För att analysera påverkan av olika parametrar på de ekonomiska mätvärdena, har en osäkerhetsanalys gjort. Den största påverkan på resultatet syntes av förändringar av investeringskostnader och ledtider. Ovanstående resultat tyder på, med transmissionslösningar enligt modellerna i detta arbete, att brytpunkten där en HVDC-lösning är mer lönsam än en HVAC-lösning inte än är passerad vid ett avstånd på 125 km till anslutningspunkten. Med en fortfarande väldigt ung teknik för HVDC, kan den ständigt utvecklande tekniken i framtiden betyda kortare ledtider och en lägre investeringskostnad för en HVDC-lösning och möjligheten att vara en mer lönsam lösning. Komplikationer med en HVAC-lösning pga den extra landkabeln, som större landanvändning och med kabeltillverkningen som en flaskhals, kan ändå göra en HVDC-lösning mer praktisk.

Detta arbete behandlar mätmetoder för detektering av partiella urladdningar (kallad PD från engelska benämningen Partial Disharge) i HVAC- och HVDC- högspänningskablar. HVAC står för ”High Voltage Alternative Current” och HVDC står för ”High Voltage Direct Current”. När en viss hög spänningsnivå överskrids i kabeln uppstår olika typer av partiella urladdningar och dessa skadar kabelns isoleringsmaterial. PD- intensiteten ökar med tiden och om de tillåts fortlöpa under en längre tid kan de med tiden helt bryta ned isoleringsmaterialet vilket leder till ett fel. Mätmetoder måste föras fram och mätinstrument måste tas i bruk för att detekterapartiella urladdningar. Målet med detta är att lokalisera eventuella skador som orsakats av partiella urladdningar, uppskatta skadorna och förhindra en total nedbrytning genom övervakning. Ett av de viktigaste målen för industri samt forskning har varit möjligheten att kunna ta fram mätinstrument som kan detektera PD i kablar och klassificera (”bedöma”) vad det är för typer av partiella urladdningar som är aktiva i kablarna. I detta arbete redovisas hur klassificering av olika typer av partiella urladdningar går till för både HVDC- och HVAC- kablar. Vidare har en mätmetod för detektering av partiella urladdningar studerats teoretiskt och testats experimentellt. Metoden är välbeprövad och kallas ”den vandrande vågmetoden”. För detta uppdrag fanns det ingen möjlighet att mäta på högspänningskablar eftersom behörighet och skyddat miljö saknats. Mätmetoden har istället prövats på en RG58- koaxialkabel eftersom att koaxialkabelns karaktäristik är jämförbar med en högspänningskabels karaktäristik. En försöksmätning med en strömsensor har också utförts eftersom att dessa normalt används i praktiken för detektering av partiella urladdningar.

This thesis attempts to investigate the performance of various reactive power compensation devices, examine the mechanism of reactive power compensation for HVDC systems, and develop guidelines for the design of reactive power compensation schemes for HVDC systems. The capabilities of various reactive power compensators to enhance power system stability are compared in both steady and transient states. An understanding of the capabilities of these compensators provides a basis for further investigation of their performance in HVDC systems. The reactive power requirements of HVDC converters are studied. The voltage dependencies of the HVDC converters at different control modes are derived, which allow for predictions of how HVDC converters impact AC system voltage stability. The transient performance of reactive power compensation options for HVDC Systems is studied by comparing their behavior during DC fault recovery, Temporary Overvoltage (TOV), and commutation failure. How to quantify the system strength when reactive compensators are connected to the converter bus is investigated. A new series of indices are developed based on the Apparent Short Circuit Ratio Increase (ASCRI). The inertia of a synchronous condenser and its impact on the frequency stability of an AC/DC system are discussed. By modelling the inverter side AC system in greater detail, the frequency stability and rotor angle stability following fault transients is examined based on time domain simulation. Finally, a guideline for designing dynamic reactive power compensation for HVDC systems is proposed.

This thesis attempts to investigate the performance of various reactive power compensation devices, examine the mechanism of reactive power compensation for HVDC systems, and develop guidelines for the design of reactive power compensation schemes for HVDC systems. The capabilities of various reactive power compensators to enhance power system stability are compared in both steady and transient states. An understanding of the capabilities of these compensators provides a basis for further investigation of their performance in HVDC systems. The reactive power requirements of HVDC converters are studied. The voltage dependencies of the HVDC converters at different control modes are derived, which allow for predictions of how HVDC converters impact AC system voltage stability. The transient performance of reactive power compensation options for HVDC Systems is studied by comparing their behavior during DC fault recovery, Temporary Overvoltage (TOV), and commutation failure. How to quantify the system strength when reactive compensators are connected to the converter bus is investigated. A new series of indices are developed based on the Apparent Short Circuit Ratio Increase (ASCRI). The inertia of a synchronous condenser and its impact on the frequency stability of an AC/DC system are discussed. By modelling the inverter side AC system in greater detail, the frequency stability and rotor angle stability following fault transients is examined based on time domain simulation. Finally, a guideline for designing dynamic reactive power compensation for HVDC systems is proposed.

Voltage-Sourced Converter High Voltage Direct Current (VSC-HVDC) transmission technology has received great interest and experienced rapid development worldwide because of its compact size, ability to connect two asynchronous AC systems and ability to connect to weak AC grids. It is expected that VSC-HVDC will play a significant role in future power transmission networks. Multi-Terminal Direct Current (MTDC) networks are even being established based on VSC-HVDC and these have great potential to support conventional AC transmission networks. However, such DC networks are vulnerable to any DC side short-circuit fault. DC protection issues must be tackled to enable the development of MTDC networks. This thesis conducts some of the underpinning research for such DC protection studies. As a first step to conduct the protection study, a detailed four-terminal VSC-HVDC system is developed in PSCAD/EMTDC, which consists of both two-level converters and MMC devices. Based on this high fidelity four-terminal system model, a thorough analysis is conducted for the two-level converter and the MMC systems under different fault scenarios. Based on this, a basic understanding of the converter systems' natural responses to these fault scenarios is obtained. Apart from using a DC circuit breaker to isolate a DC fault, there may be other devices which could potentially be used for DC protection. After the fault analysis, a study is conducted to search for any other DC protection equipment which could help the DC breaker isolate a DC fault. Different types of fault current limiters (FCLs) are reviewed and compared. It is found that the resistive type superconducting FCL (SCFCL) has the potential to be usefully employed for DC protection. Next, a DC fault detection and location strategy study is performed. This thesis conducts a detailed study of different DC fault detection and location strategies using a much higher fidelity model than previous studies. After reviewing different fault detection methodologies, it is found that wavelet transforms presently might be the best option for DC protection. The continuous wavelet transform (CWT) is then extensively tested under different DC faults and transient scenarios to prove its robustness, as this method has not been extensively studied in the previous literature. In the end, by using the CWT and placing the SCFCLs in series with DC circuit breakers, the performance of the SCFCLs under a DC side pole-to-pole fault is examined. This study shows that the SCFCL can help reduce the fault current seen by a DC breaker. In the end, a DC system fault recovery study is performed. Different methods are proposed and studied to examine the impact they have on the converter system's DC fault recovery process. A novel bump-less control is proposed to help the system achieve a good fault recovery response.

Recentes avanços em relação aos dispositivos semicondutores utilizados no processo de conversão CA/CC levaram à aplicação de conversores fonte de tensão, do inglês Voltage Source Converter (VSC), na transmissão de energia elétrica em altas tensões e corrente contínua, do inglês High Voltage Direct Current (HVDC). Uma das vantagens da utilização de VSCs é simplificr o processo de criação de redes HVDC com múltiplos terminais, identificadas pela sigla em inglês Multi-terminal HVDC (MTDC). Entretanto, a severidade das faltas em linhas CC, aliada à fragilidade dos conversores, exige a utilização de algoritmos capazes de identificar corretamente a ocorrência de faltas em um reduzido intervalo de tempo. Neste sentido, este trabalho tem por objetivo a elaboração de uma nova metodologia de proteção que possa ser aplicada na proteção primária de sistemas HVDC, especialmente para redes MTDC. Para tanto, foi elaborado um modelo detalhado de rede MTDC com três terminais e, a partir dos dados obtidos por meio de extensivas simulações de falta, foram identificadas características dos sinais de corrente na linha CC capazes de auxiliar na proteção da rede. Pela utilização da Transformada wavelet, análise de componentes principais e sistemas Genético-Fuzzy, foi possível a elaboração de um algoritmo de proteção sem comunicação, rápido, confiável e seletivo para utilização em redes MTDC. Adicionalmente, foi realizada a implementação em hardware do algoritmo proposto, evidenciando sua aplicabilidade em sistemas reais. A metodologia proposta foi capaz de garantir seletividade, confiabilidade e velocidade de atuação ao sistema de proteção, identificando corretamente faltas nos condutores CC em menos de 1,5 ms. / Recent progress regarding semiconductor devices used in AC/DC conversion led to the use of Voltage Source Converters (VSC) in High Voltage Direct Current (HVDC) power transmission systems. AN advantage of using VSCs it to simplify the creation of Multi-terminal HVDC (MTDC) networks. However, the severity of DC faults, combined with the converters vulnerability, requests the use of algorithms able to correctly identify fault occurrences in a short period of time. Therefore, this work aims to elaborate a new primary protection methodology that could be applied to HVDC systems, especially in MTDC networks. For this purpose, a detailed three terminals MTDC network has been modeled and, through extensive computational faulty simulations, DC current characteristics that are able to assist network protection methods were identified. By means of the wavelet transform, principal component analysis and genetic fuzzy systems, it was possible to develop a fast, reliable an selective non-unit protection for MTDC grids. Moreover, the proposed algorithm was implemented in hardware, emphasizing its applicability in actual systems. The proposed methodology was able to ensure selectivity, reliability and speed of operation, correctly identifying DC faults in less than 1.5 ms.

Recentes avanços em relação aos dispositivos semicondutores utilizados no processo de conversão CA/CC levaram à aplicação de conversores fonte de tensão, do inglês Voltage Source Converter (VSC), na transmissão de energia elétrica em altas tensões e corrente contínua, do inglês High Voltage Direct Current (HVDC). Uma das vantagens da utilização de VSCs é simplificr o processo de criação de redes HVDC com múltiplos terminais, identificadas pela sigla em inglês Multi-terminal HVDC (MTDC). Entretanto, a severidade das faltas em linhas CC, aliada à fragilidade dos conversores, exige a utilização de algoritmos capazes de identificar corretamente a ocorrência de faltas em um reduzido intervalo de tempo. Neste sentido, este trabalho tem por objetivo a elaboração de uma nova metodologia de proteção que possa ser aplicada na proteção primária de sistemas HVDC, especialmente para redes MTDC. Para tanto, foi elaborado um modelo detalhado de rede MTDC com três terminais e, a partir dos dados obtidos por meio de extensivas simulações de falta, foram identificadas características dos sinais de corrente na linha CC capazes de auxiliar na proteção da rede. Pela utilização da Transformada wavelet, análise de componentes principais e sistemas Genético-Fuzzy, foi possível a elaboração de um algoritmo de proteção sem comunicação, rápido, confiável e seletivo para utilização em redes MTDC. Adicionalmente, foi realizada a implementação em hardware do algoritmo proposto, evidenciando sua aplicabilidade em sistemas reais. A metodologia proposta foi capaz de garantir seletividade, confiabilidade e velocidade de atuação ao sistema de proteção, identificando corretamente faltas nos condutores CC em menos de 1,5 ms. / Recent progress regarding semiconductor devices used in AC/DC conversion led to the use of Voltage Source Converters (VSC) in High Voltage Direct Current (HVDC) power transmission systems. AN advantage of using VSCs it to simplify the creation of Multi-terminal HVDC (MTDC) networks. However, the severity of DC faults, combined with the converters vulnerability, requests the use of algorithms able to correctly identify fault occurrences in a short period of time. Therefore, this work aims to elaborate a new primary protection methodology that could be applied to HVDC systems, especially in MTDC networks. For this purpose, a detailed three terminals MTDC network has been modeled and, through extensive computational faulty simulations, DC current characteristics that are able to assist network protection methods were identified. By means of the wavelet transform, principal component analysis and genetic fuzzy systems, it was possible to develop a fast, reliable an selective non-unit protection for MTDC grids. Moreover, the proposed algorithm was implemented in hardware, emphasizing its applicability in actual systems. The proposed methodology was able to ensure selectivity, reliability and speed of operation, correctly identifying DC faults in less than 1.5 ms.