A power system will inevitably be subjected to transient overvoltages, most often produced by switching operations and lightning strikes. These transient overvoltages may harm sensitive equipment without sufficient protection, but it is not an easy task to predict what transients may occur or how they affect the system. A fault occurred in a low voltage system at Forsmark 2012 and overvoltages from lightning were concluded as the probable cause. The three aims with this thesis are to develop and test a model of the power system at Forsmark 3, analyse the transient behaviour of the system when subjected to lightning surges and identify critical parameters, and lastly to provide a tool for investigating the protection of the system. The modelling and simulations were performed in the freely available simulation program LT Spice. Challenges and difficulties have been to obtain parameter values for components that are relevant for the high frequencies produced by the short rise time in lightning surges. The main conclusions are: the maximum current in the lightning discharge has the largest impact on the system voltages and surges in the external grid can produce standing waves in the system, causing significant overvoltages even in low voltage systems. The simulations produced larger overvoltages than would occur in a real system, which also would dampen the high frequency signals to a larger extent. Further work to obtain models better suited for high frequency simulation is suggested.

The Nordic synchronous power system must continuously be in balance with respect to the electricity produced and consumed to function optimally. The system frequency is a good measure of the actual balance. The Swedish Transmission System Operator, Svenska kraftnät (Svk), is responsible for making sure such a balance is achieved. This study was conducted with the overall aim of investigating how Svk works in practice to balance the Nordic synchronous power system. This includes describing the main aspect of the power system structure, the existing methods regarding automatic and manual frequency control, the main operating systems as well as the tools and methods used by Svk. The study also examines how operational plans and forecasts available through the Nordic Operational Information System (NOIS) can be used to predict at what operational situations the risk of unwanted frequency deviations will be higher than normal. Furthermore, the study also investigates the possibility to conduct frequency regulation in a proactive manner based on these correlations. This was accomplished mainly through a literature review and interviews. The results shows that frequency deviations where the frequency is outside the interval 49.9–50.1 Hz often occur in correlation with the specific time at which the planned system balance quickly changes. Frequency simulations show that proactive regulation, which aims to reduce the planned imbalance, appears to also reduce frequency deviations. Further studies should however be conducted to study such frequency correlations in depth, and to investigate how proactive regulatory practices can be implemented in the most optimal manner.

Oscillations in power systems are a common problem that cause wear and decreased electric power quality. Hence, it is of great importance to detect when oscillations occur and to analyze which machines that are involved. This master thesis was a pre-study to implement an application that can detect and analyze power system oscillations in the real-time interactive simulator ARISTO, which is a simulator that the authority responsible for the transmission system of Sweden, Svenska kraftnät, develops. The working process included a literature study which resulted in several methods that were selected for deeper analysis. The theory of oscillation detection is called spectral analysis. To analyze which machines that are involved in the oscillations, phase shift calculations are required. The study showed that the methods ESPRIT (estimation of signal parameters via rotational invariant techniques) and Capon performed the most accurate results in the spectral analysis. Among the selected methods for phase shift calculations, the discrete Fourier transform obtained the most accurate results. Among the scientific reports and publications that discussed power system oscillations, the method GPOF (generalized pencil-of-function method) was most frequently used. The reason for this depends on the method's unique property to calculate damping, which can be a crucial aspect to analyze in power systems. This master thesis has examined several other methods that yield more accurate results in terms of spectral analysis and phase shift calculations. However, in contrast to GPOF, those methods cannot calculate the damping of the system which can be a substantial drawback.

In view of the potential threat climate change poses around the globe, countries are pledging to emission reductions to alleviate its consequences. As part of the Paris Agreement, Indonesia set sustainability targets for the energy sector, where one of the biggest emitters is the national power system. Although the country set an ambitious strategy for reducing emissions from electricity generation, it is still dominated by coal consumption and is currently not on track towards meeting its sustainability targets. Policy measures and financial schemes stand as one of the most powerful ways to rapidly reduce emissions in the power system. This paper aims to propose policy measures and financial schemes for the Indonesian decarbonisation scene by analysing its key inefficiencies, the available international policy options and the national power system’s future projections.International examples prove to be useful implementations of the myriad of options when it comes to measures for the power sector. In terms of policy measures, Indonesia could make use of a liberalisation process to induce more competitiveness in the sector and drive down prices, while allowing easier access for private entities to enter the market. A CO2 tax or cap-and-trade system stand as the most effective standalone way to rapidly reduce emissions in the country, as it targets the most emitting sectors of the power system. Incentivising schemes like Feed-in-Tariff or net-metering foster generation technologies which may not be cost-competitive yet and do not have a sizeable role in the power mix now, but could play a big role in lowering emissions in the power system. These measures, however, pose the threat of raising the retail electricity price and ultimately hinder Indonesia’s ability to further develop economically.Financial schemes have the ultimate goal of facilitating access to finance for sustainable generation technologies and lowering risk for project developers. Attracting private and foreign investment is critical for the development of the power system in Indonesia. Using forms of lending like on-lending structure or loan syndication would promote larger spending in the industry while minimising risk. Further use of guarantees would attract a wider range of investments into generation technologies, as they would protect lending entities from potential risks when partaking in projects of this nature. Finally, the country would benefit greatly from the introduction of innovative financial schemes and business plans for the domestic solar energy industry, allowing Indonesia to tap into the great potential it has to disrupt the current state of affairs of the power system. / Med tanke på det potentiella hot som klimatförändringen utgör runt om i världen, lovar länder att minska utsläppen för att mildra konsekvenserna. Som en del av Parisavtalet satte Indonesien upp hållbarhetsmål för energisektorn, där en av de största utsläpparna är det nationella kraftsystemet. Trots att landet satte upp en ambitiös strategi för att minska utsläppen från elproduktion domineras det fortfarande av kolkonsumtion och är för närvarande inte på väg att nå sina hållbarhetsmål. Politiska åtgärder och finansiella system är ett av de mest kraftfulla sätten att snabbt minska utsläppen i kraftsystemet. Det här dokumentet syftar till att föreslå politiska åtgärder och finansiella system för den indonesiska avkarboniseringsscenen genom att analysera dess viktigaste ineffektivitet, de tillgängliga internationella politiska alternativen och det nationella kraftsystemets framtida prognoser.Internationella exempel visar sig vara användbara implementeringar av de otaliga alternativen när det gäller åtgärder för kraftsektorn. När det gäller politiska åtgärder skulle Indonesien kunna använda sig av en liberaliseringsprocess för att skapa mer konkurrenskraft inom sektorn och pressa ner priserna, samtidigt som privata enheter lättare kan komma in på marknaden. En CO2-skatt eller ’cap-and-trade’-system är det mest effektiva fristående sättet att snabbt minska utsläppen i landet, eftersom det är inriktat på de mest utsläppande sektorerna i kraftsystemet. Incitamentsystem som ’Feed-in-Tariff’ eller ’net-metering’ främja generationstekniker som kanske inte är kostnadskonkurrenskraftiga ännu och som inte spelar en betydande roll i kraftmixen nu, men som kan spela en stor roll för att minska utsläppen i kraftsystemet. Dessa åtgärder utgör dock ett hot om att höja detaljhandelns elpris och i slutändan hindra Indonesiens förmåga att utvecklas ekonomiskt.Finansiella system har det yttersta målet att underlätta tillgången till finansiering för hållbara generationsteknologier och minska risken för projektutvecklare. Att locka till sig privata och utländska investeringar är avgörande för utvecklingen av kraftsystemet i Indonesien. Att använda former av utlåning som ’on-lending’-struktur eller lånesyndikering skulle främja större utgifter i branschen samtidigt som risken minimeras. Ytterligare användning av garantier skulle locka till sig ett bredare utbud av investeringar i produktionsteknik, eftersom de skulle skydda långivande enheter från potentiella risker när de deltar i projekt av denna karaktär. Slutligen skulle landet dra stor nytta av införandet av innovativa finansiella system och affärsplaner för den inhemska solenergiindustrin, vilket gör att Indonesien kan utnyttja den stora potential det har att störa det nuvarande tillståndet i kraftsystemet.

The requirement for reliability and security in power system is increasing every passingday with the increase in complexity of the power system. To ensure highest level ofreliability, protection relays have to receive correct measurements. One of the most importantmeasurement that is needed by a relay as an input is current. However, currentmeasurements received from current transformer (CT) can become inaccurate due to aphenomenon called CT saturation. This Master's thesis objective is to build a novel algorithmfor the protection relays to detect CT saturation. The algorithm should be basedon a stand-alone method that is able to detect CT saturation within 1-2 ms for a samplingfrequency of 4 kHz.This project comprises a study of the background about CTs and CT saturation. Therelated work done to detect CT saturation is also studied. Later, major existing methodsused in the industry to deal with CT saturation are investigated in detail and modelledin Simulink. A novel method is proposed to exclusively detect CT saturation, keeping inmind the strict requirements, set in the beginning of the project. The proposed methodis implemented in Simulink and HiDraw(in-house software in ABB to create C code forprotection functions). Lastly, the proposed method and the existing methods are testedin Simulink for more than 1300 test cases based on dierent power system conditions,IEC-60255-187-1 and real current recordings from relays.The results of the tests showed that the proposed method successfully detect CT saturationand was better than the existing methods in terms of speed and accuracy. It was concludedthat the proposed method successfully detects CT saturation and hence, can be used withany protection function in a relay where there is a need to detect CT saturation. / Kraven pa tillforlitlighet och sakerhet i kraftsystem okar varje dag i samband med allthogre komplexitet i kraftsystemet. For att uppna den hogsta nivan av tillforlitlighetbehover relaskydden ta emot korrekta matvarden fran kraftsystemet. Ett av de mest viktigamatvardena som behovs for ett relaskydd ar strom. Emellertid, kan matvarden franen stromtransformator bli felaktiga pagrund av ett fenomen som kallas stromtransformatormattning. Malet for detta examensarbete ar att skapa en ny algoritm for relaskydd somdetekterar stromtransformatormattning. Algoritmen ska baseras pa en fristaende metodsom klarar av att detektera stromtransformatormattning inom 1-2 ms for en samplingsfrekvenspa 4 kHz.Detta projekt omfattar, forutom en studie av hur stromtransformatormattning kan detekteras,aven en bakgrundsstudie om stromtransformatorer och stromtransformatormattning.Viktigare existerande metoder, som anvands i industrin for att detektera stromtransformatormattning, undersoks i detalj och modelleras i Simulink. En ny metod foreslas foratt exklusivt detektera stromtransformatormattning med hansyn till de strikta krav somfaststalldes i borjan av projektet. Den foreslagna metoden ar implementerad i Simulinkoch i HiDraw (en intern ABB mjukvara for att skapa C-kod for skyddsfunktioner). Slutligenar den nya foreslagna metoden och de existerande metoderna testade i Simulinkmed er an 1300 testfall baserade paolika forhallanden i kraftsystemet, IEC-60255-187-1standarden och med verkliga inspelningar av strom gjorda av relaskydd i drift.Resultatet av testerna visar att den nya foreslagna metoden framgangsrikt detekterarstromtransformatormattning och gor det battre an existerande metoder med avseende pasnabbhet och noggrannhet. Det konstaterades att den nya foreslagna metoden framgangsriktdetekterar stromtransformatormattning och darfor kan anvandas for vilken relaskyddsfunktionsom helst i ett relaskydd dar behovet av att detektera stromtransformatormattningnns.

It is well known that a crucial factor determining the transmission capacity with regards to the voltage stability limit of a stressed power system, is the limitation of field and stator current of overexcited generators. These limiters are commonly referred to as overexcitation limiters (OEL) and stator current limiters (SCL). This thesis investigates the representation of overexcitation limiters and stator current limiters in the Swedish power system and how they could be implemented in the power system simulator PSS/E. Dynamic simulations in PSS/E are performed to investigate and validate differences between model types and parameters. The thesis compares different types of limiters and presents a representative set of parameters based on the documentation and implements them into PSS/E. The result from the study is a recommendation to migrate to limiter model MAXEX2 for generators without stator current limiters. The MAXEX2 model uses an inverse time characteristic which is becoming more common in installed voltage regulators. This feature makes it possible for extended use of generators before limiter action is applied. The MAXEX2 model is also possible to tune into a fixed time delay to represent older types of limiters.

Svenska kraftnät is responsible for the operation and development of the Swedish transmission grid. A large number of substations in the transmission grid are approaching their expected service lifetime and must be renewed in the coming years. When renewing substations, total and partial outages in operation are often unavoidable. These outages may cause downtime for any party connected to the substation, whether it be an electricity generation or consumption. A solution is to build a new substation while the old is still in operation. Upon completion, a transfer of operation can be done from the old to the new substation. This however requires laying claim to new land, which affects the local environment and can be a time-consuming process. In an effort to minimise the need for planned outages during substation renewals, Svenska kraftnät wants to investigate techniques for bypassing substations while also allowing the connected party to remain in operation. This thesis has evaluated two differing techniques that allow for substations and any connecting parties to remain mostly in full operation during station renewals. One technique involves constructing a temporary switchgear compartment using air-insulated apparatuses to bypass the station and connect the third party. The other technique uses a mobile gas-insulated switchgear to achieve essentially the same arrangement.  After studying cases where either technique has been employed in station renewals, it was concluded that a solution utilising air-insulated apparatuses is better in several aspects. While the two technologies achieved similar results, the air-insulated technology did so at a lower cost, offered more flexibility, and was easier to integrate into pre-existing stations. Furthermore, the gas-insulated technology required more specialised competence, possibly limiting the availability of qualified contractors. It was concluded that bypassing stations during renewal projects could provide several benefits in terms of less downtime and economical losses, while also minimising the need to claim new land. In order to optimise the method a supply of air-insulated apparatuses and construction materials necessary for the bypassing of a station should be established by Svenska kraftnät. This, mostly reusable, equipment would then be on stand-by, ready to use in future station renewals.

This thesis explores the use of machine learning as a load classifier in a distribution grid based on the daily consumption behaviour of roughly 1600 loads spread throughout the areas Bromma, Hässelby and Vällingby in Stockholm, Sweden. Two common unsupervised learning methods were used for this, K-means clustering and hierarchical agglomerative clustering (HAC), the performance of which was analysed with different input data sets and parameters. K-means and HAC were unfortunately difficult to compare and there were also some difficulties in finding a suitable number of clusters K with the used input data. This issue was resolved by evaluating the clustering outcome with custom loss function MSE-tot that compared created clusters with subsequent assignment of new data. The loss function MSE-tot indicates that K-means is more suitable than HAC in this particular clustering setup. To investigate how the obtained clusters could be used in practice, two K-means clustering models were also used to perform some cluster-specific peak load predictions. These predictions were done using unitless load profiles created from the mean properties of each cluster and dimensioned using load specific parameters. The developed models had a mean relative error of approximately 8-19 % per load, depending on the prediction method and which of the two clustering models that was used. This result is quite promising, especially since deviations above 20 % were not uncommon in previous work. The models gave poor predictions for some clusters, however, which indicates that the models may not be suitable to use on all kinds of load data in its current form. One suggestion for how to further improve the predictions is to add more explanatory variables, for example the temperature dependence. The result of the developed models were also compared to the conventionally used Velander's formula, which makes predictions based on the loads' facility-type and annual electricity consumption. Velander's formula generally performed worse than the developed methods, only reaching a mean relative error of 40-43 % per load. One likely reason for this is that the used database had poor facility label quality, which is essential for obtaining correct constants in Velander's formula.

The use of electricity from Distributed Energy Resources like wind and solar powerwill impact the performance of the electricity network and this sets a limit to theamount of such renewables that can be connected. Investment in energy storage andcommunication technologies enables more renewables by operating the networkcloser to its limits. Electricity networks using such novel techniques are referred toas “Smart Grids”. Under favourable conditions the use of these techniques is analternative to traditional network planning like replacement of transformers orconstruction of new power line.The Hosting Capacity is an objective metric to determine the limit of an electricitynetwork to integrate new consumption or production. The goal is to create greatercomparability and transparency, thereby improving the factual base of discussionsbetween network operators and owners of Distributed Energy Resources on thequantity and type of generation that can be connected to a network. This thesisextends the Hosting Capacity method to the application of storage and curtailmentand develops additional metrics such as the Hosting Capacity Coefficient.The research shows how the different intermittency of renewables and consumptionaffect the Hosting Capacity. Several case studies using real production andconsumption measurements are presented. Focus is on how the permitted amountof renewables can be extended by means of storage, curtailment and advanceddistributed protection and control schemes. / Användningen av el från förnyelsebara energikällor som vind och sol kommer att påverka elnätet, som sätter en gräns för hur mycket distribuerad energiproduktion som kan anslutas. Investeringar i storskalig energilager och användning av modern kommunikationsteknologi gör det möjligt att öka andelen förnyelsebarenergi genom att nätet kan drivas närmare sina gränser. Elnät med sådana nya tekniker kallas ofta för ”Smarta Elnät". Implementering av sådana smarta elnät kan vara ett alternativ till traditionell nätplanering och åtgärder som utbyte av transformatorer eller konstruktion av nya kraftledningen.Nätets acceptansgräns är ett objektivt mått för att bestämma gränsen för nätets förmåga att integrera ny förbrukning eller produktion. Målet är att skapa större transparens och bidra till ett bättre faktaunderlag i diskussioner mellan nätoperatörer och ägare av distribuerade energiresurser. Denna avhandling utökar acceptansgränsmetoden för tillämpning med energilager och produktions nedstyrning och utvecklar ytterligare begrepp så som acceptansgränsen koefficienten.Forskningen visar hur varierbarheten hos olika förnyelsebara energikällor samverkar med förbrukningen och påverkar nätets acceptansgräns. Flera fallstudier från verkliga elnät och med uppmätt produktion och konsumtion presenteras. Fokus är på hur den tillåtna mängden förnyelsebara energikällor kan ökas med hjälp av energilagring, kontrollerad produktionsnedstyrning och med avancerad distribuerade skydd och kontroll applikationer. / Nicholas Etherden works at STRI AB (www.stri.se) in Gothenburg, Sweden. When he is not pursuing his half-time PhD studies he works as a specialist consultant in the field of Power Utility Automation, specialising on the IEC 61850 standard for power utility automation (today widely used in substations as well as some wind parks, hydro plants and DER and Smart Grid applications such as vehicle-to-grid integration). The author of this thesis received his Master of Science in Engineering Physics from Uppsala University 2000. Side tracks during his engineering studies included studies in theoretical philosophy, chemistry, ecology and environmental sciences as well as chairing the Swedish student committee of the Pugwash Conferences on Science and Worlds Affairs and later board member of the International Network of Engineers and of Scientists for Global Responsibility (INES) and chair of Swedish Scientists and Engineers Against Nuclear Arms. He has been a trainee at ABB in Västerås Sweden and spent six years as developer and team leader for the application development of a new relay protection family (ABB IED 670 series). In parallel to his professional work he studied power system engineering at Mälardalens University and travelled to all continents of the world. Since 2008 he is responsible for the STRI IEC 61850 Independent Interoperability Laboratory and a member of IEC Technical Committee 57 working group 10 "Power system communication and associated data models” and UCA/IEC 61850 User group testing subcommittee. He is co-author of IEC 61850-1 and main contributor to “Technical Report on Functional Test of IEC 61850 systems” and has held over 25 hands-on courses around the world on IEC 61850 “Communication networks and systems for power utility automation”. / SmartGrid Energilager

Decision problems in operation and planning of power systems often rely on large-scale models and data sets. Lack of historical power flow data due to regulatory restrictions often limits researchers to study the system with aggregated network models. Aggregated data from the electricity market operators (Nordpool in the Nordics) and the Transmission System Operator (TSO) (from ENTSO-E) are openly available, and can be used to study the power flow and exchanges between different regions but do not directly provide information about intra-region flows. This project builds upon the Nordic 490 system, a previously built model of the Nordic power system. The main objective of this work is to improve the existing open source power flow model of the Nordic power system, in order to become in turn available for the multi-energy modelling and simulation software Spine. The N490 model generates a model of the Nordic power system consisting of various nodes/buses which represent substations at different voltage levels. Then, it distributes the aggregated production, consumption and power exchange data from Nordpool to the various buses. In this project, different possible improvements are evaluated for the model, aiming at estimating a set of network parameters that minimize the errors between the calculated inter-region flows and the ones from the open data repositories. The different improvements which are evaluated are the following. Firstly, the load distribution is modified and reassigned to match the regional electricity consumption. The generators and wind farms are then reallocated to different bus based on their bidding region and proximity to the bus. The databases are improved and the power balance relation modified. Transmission line parameters are then investigated, first to standard recommended values and then by solving an optimisation problem formulated to extract the parameters from the market data. Finally, the model is also tested with wind and solar generation modelled as a generator rather than as a negative load. / Beslutsproblem gällande drift och planering av kraftsystemet baseras ofta på storskaliga modeller och datamängder. Bristen på historiska data gällande effektflöden beror på säkerhetsrestriktioner vilket begränsar forskare till att enbart studera aggregerade nätverksmodeller. Det finns tillgängliga aggregerade data från den nordiska elmarknadsplatsen Nordpool och organisationen ENTSO-E som kan användas för att studera effektflöden mellan olika regioner, dock finns det inte direkta data för flöden inom regionerna. Det här projektet bygger på det nordiska 490-systemet, en tidigare byggd modell av det nordiska kraftsystemet. Huvudsyftet med detta arbete är att förbättra den existerande effektflödesmodellen av det nordiska kraftsystemet, för att i sin tur bli tillgänglig för multienergimodelleringar och simuleringsprogramvaran Spine. N490-modellen genererar en modell för det nordiska kraftsystemet som innehåller olika noder som presenterar ställverk med olika spänningsnivåer och modellen ger också aggregerade data för produktion, konsumtion och effektutbyte mellan de olika noderna från Nordpool. I detta projekt utvärderades olika möjliga förbättringar för modellen som syftar till att uppskatta nätverkets parametrar som kan minimera felen mellan beräkningar av flöde inom regionen och data från öppna datalagringskällor. Följande förbättringar gjordes: Först har lastens fördelning modifierats och ändrats för att matcha den regionala elkonsumtionen. Generatorer och vindkraftsparker allokerades till olika noder baserad på elhandelsområden och närhet till noderna. Databasen förbättrades för att erhålla en bättre effektbalans per område. Kraftledningarnas parametrar ändrades först till rekommenderade standardvärden, vilka sedan förbättrades genom att formulera ett optimeringsproblem för att extrahera parametrarna från markandsdata. Slutligen testades modellen genom att presentera vind- och sol-produktion som generatorer istället för som negativ förbrukning.