With the increase in energy demand, and therefore, in energy generation, new high voltage direct current export links and overseas connections of large amounts of power are being introduced. These new links are subject to possible malfunction or failures, which may causea substantial frequency increase in the power system. The need of restorative measurements for over-frequencies is then necessary to regulate this probable imbalance. A new ancillary market product for frequency down-regulation is then being introduced. The Frequency Containment Reserves – Disturbances for downregulation is a service used to handle overfrequencies in the power system. The ancillary market has been historically dominated by hydro, however, with the local plans that set goals to reach 100 % renewable electricity production by the year 2040, ineluctably solar and wind will be occupying a large share of the energy generation, thus the amount of renewable resources to provide ancillary services will increase as well. Nevertheless, requirements to participate in the ancillary market are strict, and renewables face challenges such as unpredictable power generation, making forecasting tools essential for their participation. In this thesis, through the literature and research of an emerging field, the synthetic solar irradiance, an algorithm with adjusted parameters is replicated with the objective of downscaling solar irradiance presumptive-forecasts from 15-minute intervals to 1-minute intervals. An updated clear-sky model, and weather categorization method is used as input to improve the results of the synthesis. A year evaluation with a method to detect irradiance dips is used to evaluate the algorithm. The result of this study contributes with positive outputs. The introduction of this secondary tool to estimate higher resolution irradiance variability data proves to obtain, at 57 % of the point-to-point critical cases examination, an over-estimation of the measured irradiance. This, in turn, grants the actors that participate with solar power the option to adjust the bidding strategy process in the ancillary market accordingly to the output of this tool.

This diploma thesis deals with the simulation of various operating situations in the network with many non-rotating power sources, especially focusing on frequency stability. Non-rotating power source is any source that delivers its power to the grid via power electronics. The first part of the thesis describes power system of the Czech Republic and its future development. In the next section, there is short description of today's most common non-rotating power sources. The third part deals with power regulation of the frequency in the grid. The following part describes the models of electrical circuit created for use in PSCAD simulations. The penultimate part describes simulations performed on a model of the network with many non-rotating power sources, which also includes rotating sources. The last part describes simulations performed on a model of the network that is composed of non-rotating power sources only.

Elnätet är ett komplext och viktigt system som konstruerats som en av de mest imponerande bedrifterna under ingenjörskonstens moderna tid. Det överför elektrisk energi till otaliga byggnader, industrier, skolor och hem. Och alltsammans sker konstant, varje minut av varje dag, året runt.  Grundbulten i systemet är den att en ständig balans måste råda mellan produktion och förbrukning av elektricitet. Vid obalans riskerar nämligen strömavbrott och andra oönskade företeelser att inträffa. Huruvida produktion och förbrukning av elektricitet är i nivå kan beskrivas av elnätets frekvens. Genom att övervaka elnätsfrekvensens beteende fås en överskådlig bild av elnätets status i realtid samtidigt som stödtjänster kan implementeras proaktivt för att motverka eventuella störningar. Tidigare har dessa stödtjänster främst representerats av stora aktörer såsom vattenkraftsanläggningar med enorma förutsättningar för att agera som reglerkraft. I takt med en omställning till en alltmer förnybar energipalett ökar dock behovet av ny reglerkraft. Samtidigt syns ett accelererande av installering av batterilager i bostadsrättsföreningar som energieffektiviserar. Eventuellt finns här en outforskad potential. Möjligen kan batterilager i flerbostadshus utnyttjas för frekvensreglering som en ytterligare balanskraft för elnätet. Projektarbetet syftade till att utreda potentialen kring batteristyrd mFRR-reglering från flerbostadshus. För att utvärdera den eventuella lönsamheten modellerades batteristyrningen i MATLAB. Modellen baserades främst på historiska data för upp- och nedregleringsbud från Nord Pool. I och med kravet på 1 MWh som minsta budstorlek på marknaden gjordes antagandet att vid varje regleringstillfälle reglerar batterilagret i aggregation med andra batterilager som tillsammans täcker den totala kapaciteten på 1 MWh. Modellen fungerar på så sätt att varje uppreglering föranleds av en uppladdning av batterilagret via antingen nedreglering eller spotpriser. Beroende på vilket alternativ som är mest lönsamt. Vidare gjordes antagandet att inga uppregleringsbud sker i två påföljande timmar.  Studiens mest lönsamma resultat genererades då batterilagret modellerades för att anta ett uppregleringsbud per dygn med det extra villkoret att samtliga bud som understiger 300 SEK/MWh förkastas. Vid dessa kriterium erhölls ett positivt årligt resultat om 149 100 kr från 306 battericykler.  Med endast frekvensreglering som användningsområde för batterilagret konkluderade dock studien att, det positiva resultatet till trots, ingen lönsamhet kunde uppnås. Investeringskostnaden är nämligen ännu för hög. Å andra sidan tyder teknologiska framsteg inom batterisektorn på en avtagande kostnadsutveckling. Vid år 2030 väntas nämligen priset för batterilager vara 225 USD/kWh, vilket skulle förbättra resultaten från denna rapport. / The electricity grid is a exceptionally sophisticated and crucial system which was created as one of the most impressive accomplishments in modern engineering. It transmits electrical energy to innumerable buildings, factories, schools and homes. And it can never stop. The system relies on the constant balance between generated and consumed electricity. Should an imbalance occur, it might give rise to power outages or failures in appliances on the grid. Whether or not generated and consumed electricity is at balance is determined by the grid frequency. By surveilling how the grid frequency behaves, an overview of the complete system can be obtained in real time. This is useful when it comes to deciding on whether or not to implement supporting mechanisms to counteract disturbances. The supporting mechanisms were historically represented by large facilities such as water power plants who possess great abilities of regulating the power balance on the grid. With the ongoing switch to renewables the need for more regulating power increases. At the same time installments of battery storage in energy efficient housing can be seen as accelerating. Here might be an untapped potential. What if battery storage in residential properties were utilized for frequency regulation as an additional balancing tool for the grid? The intent with this report was to outline the potential regarding power regulating via battery storages in multifamily residentials. In order to evaluate whether or not any profit could be redeemed, a battery control model was developed in MATLAB. The model was primarily based on historic data for regulating bids from Nord Pool. Since requirements on the market states that the lowest bid needs to be at least 1 MWh, the assumption was made that at every regulating occasion the battery regulates in aggregation with more batteries for a total capacity of 1 MWh. The objective of the model is to charge the battery using either spot price or down-regulating bids. Thereafter, an up-regulating bid is chosen. The assumption was made that no subsequent up-regulating bids were chosen. The most profitable optimization of the model was generated when 1 up-regulating bid was chosen per day with the additional condition that all bids under 300 SEK/MWh were rejected. At these criteria a yearly positive result of 149 100 kr was generated from 306 battery cycles. With frequency regulation as sole application for the battery storage, the study concluded that the model was not profitable. The cost of investment is yet too high. On the other hand, technological improvements will surely amount to declining prices. By year 2030 the price of battery storage may have fallen to 225 USD/kWh, which would improve the results in this study.

With global climate change as the main driver, there is an increase towards including more variable renewable energy (VRE) sources in the electricity mix. Energy production from utilizing the photovoltaic effect, or PV power, is increasing rapidly and is visioned to cover 5 – 10 % of Sweden’s electricity demand in 2040. In addition to rooftop PV production, large- scale PV production in the form of ground-mounted PV parks is gaining ground. A higher share of VRE in the power system creates new challenges as to uphold the power system stability. For a PV park owner, achieving a preferable economic outcome is also a challenge, as the variable electricity output may not match electricity demand. Therefore, combining a PV park with an energy storage, which can store the PV production energy, is seen as a favorable solution. This way, the variability of the electricity production can be reduced and the stored energy in the battery can be used for services benefitting both the PV park owner and the power grid. This study aims to explore the economic potential of combining a PV park with an energy storage. This is achieved by simulating a lithium-ion (Li-ion) battery storage combined with PV production modeled after a 3.5 MW PV park located in Fyrislund, Uppsala. Five cases with individually differing approaches are simulated, exploring how so-called service stacking can be applied with a battery. The investigated services included in the cases are 1) lowering the cost of connecting the PV park to the power grid, 2) lowering the cost of feeding in energy to the power grid, 3) increasing the revenue of selling electricity on the Nord Pool spot market, 4) increasing the revenue by performing energy arbitrage, 5) increasing the revenue by participating in the primary frequency regulating markets to help stabilize the 50 Hz grid frequency. The cases are evaluated by calculating the net present value (NPV) of the system over 10 years with an annual discount rate of 5 %. Battery capacities ranging from 0.1 MWh/0.1 MW to 8 MWh/2 MW are tested. The system configuration achieving the highest NPV occurs when all services are performed, and a 0.13 MWh/0.1 MW battery is used. This NPV is also higher than the NPV when not including a battery in the system. Conclusions include that the spot price impacts the choice of battery capacity to a high extent and that the battery investment cost motivates using a smaller-sized battery.

Le placement de production (UC pour unit commitment) est une famille de problèmes d'optimisation qui déterminent l’état et la puissance de consigne des groupes de production pour satisfaire la demande électrique à moindre coût. Traditionnellement, une contrainte de sûreté détermine un certain volume de capacité raccordée disponible, appelé la réserve, destinée à gérer l'incertitude. Néanmoins, dans les petits systèmes la contrainte de réserve fixe peut entraîner dans certains cas une violation du critère N-1 bien que le volume de réserve minimale soit respecté. Plus récemment, la part croissante de production variable à partir de sources renouvelables (ENR) peut conduire à des programmes d’appel qui ne garantissent plus la sûreté même dans les grands systèmes.Pour y faire face, différentes techniques d'atténuation des impacts ont été proposées telle que la révision des modèles de placement de la production pour inclure une meilleure représentation de la dynamique du système. Cette sous-famille des problèmes UC est formellement définie dans ces travaux comme le problème FCUC (frequency constrained unit commitment). Elle vise à maintenir la fréquence au-dessus d'un certain seuil, et éviter ainsi le délestage par sous-fréquence (DSF).La première partie de ces travaux identifie les défis dans la formulation du problème FCUC. D’une part, la contrainte de fréquence est fortement non-linéaire par rapport aux variables de décision du problème UC. D’autre part, elle est difficile à approcher par des fonctions analytiques. La simulation séquentielle d'un modèle UC classique et d’un modèle de réponse primaire de la fréquence est alors proposée. L’intérêt d’une formulation plus fidèle de la contrainte de sûreté est donc révélé. La deuxième partie de ces travaux étudie l'impact des ENR sur la réponse primaire de la fréquence. Le besoin de formuler des modèles de FCUC plus précis est mis en avant.La troisième partie des travaux examine le coût, les bénéfices et les limitations des modèles FCUC, basés sur des contraintes indirectes sur certains paramètres dynamiques des unités de production. Il est montré que, bien que l'application de contraintes de sécurité indirectes assure la sûreté dans certains pas horaires, l'effet inverse peut apparaître à un autre instant. Ainsi, l’efficacité des leviers dépend fortement du point de fonctionnement du système. Il en est de même pour le coût de la solution. Cette étude met en évidence la nécessité de nouvelles méthodes pour traiter correctement la contrainte sur le creux de fréquence afin d'assurer l'optimalité et efficacité de la solution.Finalement, la quatrième partie des travaux offre une nouvelle formulation du problème FCUC suivant une approche de décomposition de Bender. La décomposition de Bender sépare un problème d'optimisation avec une certaine structure en deux parties : le problème maître et le problème esclave. Dans le cas du FCUC, le problème maître propose des plans de production candidats (états des groupes) et le problème esclave assure le respect des contraintes de fréquence par le biais d'un modèle de plans sécants. Les résultats de simulation montrent que la représentation plus précise du creux de fréquence au niveau du problème esclave réduit le risque de DSF et le coût de la sécurité par rapport à d'autres modèles de FCUC. / The Unit Commitment problem (UC) is a family of optimisation models for determining the optimal short-term generation schedule to supply electric power demand with a defined risk level. The UC objective function is given by the operational costs over the optimisation horizon. The constraints include, among others, technical, operational and security limits. Traditionally, the security constraints are given by the requirement of a certain volume of on-line spare capacity, which is called the reserve and is meant to handle uncertainty, while preventing the interruption of power supply. It is commonly specified following a static reliability criterion, such as the N-1 rule.Nevertheless, in small systems the fixed, and a priori defined, reserve constraint could entail a violation of the N-1 criterion, although the reserve constraint was met. More recently, the increasing share of variable generation from renewable sources (V-RES), such as wind and solar, may lead to UC solutions that no longer ensure system security. Therefore, different impact mitigation techniques have been proposed in literature, which include the revision of UC models to provide a better representation of the system dynamics. This subfamily of UC models is formally defined in this work as the frequency constrained UC problem (FCUC), and aims to keep the frequency above a certain threshold, following pre-defined contingencies, by adding enhanced security constraints. In this work this topic is addressed in four parts.The first part identifies the main challenge of formulating the FCUC problem. Indeed, the frequency minimum, also called the frequency nadir, constraint is strongly non-linear on the decision variables of the UC model. Moreover, the behaviour of the frequency nadir regarding the binary decision variables is hard to approximate by analytical functions. Thus, a sequential simulation approach is proposed, based on a classic UC model and a reduced order model of the primary frequency response. The potential benefits of a smarter allocation of the primary reserve is revealed.The second part of this work investigates the impact of V-RES sources on the primary frequency response. The underlying processes that lead to the increase of the Under-Frequency Load Shedding (UFLS) risk are thoroughly discussed. The need of formulating more accurate FCUC models is highlighted.The third part of this work examines the cost/benefit and limitation of FCUC models based on indirect constraints over certain dynamic parameters of the generating units. A methodology is proposed that assesses the effectiveness and optimality of some existing V-RES impact mitigation techniques, such as the increase of the primary reserve requirement, the prescription of an inertia requirement, the authorisation of V-RES dispatch-down or the consideration of fast non-synchronous providers of frequency regulation services. This study showed the need for new methods to properly handle the frequency nadir constraint in order to ensure optimality, without compromising the optimisation problem’s tractability.The fourth part of this work offers a new formulation of the FCUC problem following a Bender’s decomposition approach. This method is based on the decomposition of an optimisation problem into two stages: the master and the slave problems. Here, the master problem deals with the generating unit states and the slave problem handles the frequency nadir constraints through a cutting plane model. Simulation results showed that the more accurate representation of the frequency nadir in the slave problem reduces the risk of UFLS and the security cost, with respect to other FCUC models, such as those based on inertia constraints. In addition, the optimality of the global solution is guaranteed; although the convergence of the master problem is slow, due to the well-known tailing off effect of cutting plane methods.

Integrating more renewable energy sources into the grid has caused increased instability due to the intermittency of renewable energy sources. Hence, the need for grid balancing strategies such as frequency regulation has intensified. Areim, a Nordic real estate investment company, through this thesis, aims to have an assessment conducted to estimate the environmental benefits or consequences of using their specific battery system as a participant in the Swedish frequency regulation market, using the lifecycle assessment framework. The study only considered the cradle-to-gate lifecycle scope, excluding the product disposal stage, and the impact categories used align with the Environmental Footprint assessment methodology. The functional unit is in per kilo-watthour delivered, and the batteries are expected to deliver 933 kWh of electric energy over the estimated lifetime of 15 years. The normalized carbon emissions caused by delivering 1 kWh of energy for frequency regulation using the status quo prequalified technologies primarily comprised of hydropower, combined heat and power, and battery energy storage produce 4.75 kgCO2eq. Introducing Areim's specific battery system 200 kW bid into the prequalified technologies mix by substitution produces 0.075 kgCO2eq fewer carbon emissions per kWh delivered. The sensitivity analysis further supports that Areim will yield added carbon emission savings by increasing its available prequalified re-source capacity in the market. The findings of this thesis can be used to support Areim and other companies interested in grid support services such as frequency regulation to decide whether it is beneficial to use their specific battery systems for such services from an environmental effect perspective.

The residential and commercial sector is the largest consumer of electricity in Sweden and therefore highly affected by fluctuations in electricity price. On the other hand, there is a large potential to reduce both the electricity demand and emissions from electricity generation on a national level if measures are taken within this sector. This Masters' Thesis focuses on the implementation of Solar PV panels and Lithium-ion Battery Energy Storage Systems (Li-BESS) in commercial buildings. The thesis was conducted in collaboration with Vasakronan, one of Sweden’s largest real-estate companies in terms of market share with a lot of focus on sustainability and reducing the electricity demand of their properties. The objective of this study is to determine which one out of three management strategies: Peak Shaving, Energy Arbitrage and Frequency Regulation that generate the highest revenue for a Solar PV + Li-BESS system in Vasakronan’s Lumi Property. A research gap was identified in the literature for a techno-economic model evaluating and comparing these three control strategies, where Peak Shaving and Energy Arbitrage are established, while Frequency Regulation is a new possible control strategy in commercial buildings. The techno-economic model is developed for the three control strategies using the System Advisor Model and Microsoft Excel. The model evaluates battery sizes between 0-600 kWh and uses input data on weather and electricity prices for the years 2018-2022. Quantifiable outputs in terms of Key Performance Indicators (KPIs) from the model are compared between control strategies. Net Present Value (NPV) and Internal Rate of Return (IRR) are the main economic KPIs to determine which control strategy is the most profitable. The result of the analysis showed that the Peak Shaving and Energy Arbitrage controllers does not reach a positive NPV or an IRR above the set discount rate of 6.5 % for any of the battery sizes above 120 kWh for any of the studied datasets. The main reason is that the investment cost for a battery large enough to reduce a sufficient part of the demand or generate enough revenue from the difference in electricity price, is too high. With no battery at all, or a small battery of 120 kWh, the savings in electricity cost from the Solar PV system surpasses the investment and entails a profitable result. The Frequency Regulation controller generated a positive NPV and IRR above the discount rate for all battery sizes between 240 – 600 kWh for all years except 2021, which was the year with the lowest solar radiation. The conclusion from the result for the Frequency Regulation controller is that the revenue gained from a 120-kWh battery is not enough to cover for the investment cost, but for larger batteries the revenue exceeds the investment cost over the whole modelling period. The overall conclusion from is that battery storage with a Li-BESS in commercial buildings coupled with a Solar PV system is only profitable if the battery capacity is offered on the FCR markets. In all other cases, it is preferable to have a solar PV system without battery storage from an economical point of view. However, new markets and potential business models are developing continuously which is something future studies should investigate. A more thorough analysis of the frequency regulation markets and how these will develop over time is required to further validate the result of this thesis. / Sektorn för Bostäder och Service är den största elkonsumenten i Sverige och påverkas därför starkt av fluktuationer i elpriset. Samtidigt finns en stor potential att minska elbehovet och utsläppen från elproduktion på nationell nivå om åtgärder vidtas inom denna sektor. Denna masteruppsats fokuserar på implementeringen av solclellspaneler och batterilagring med ett Lithium-jon-batteri i komersiella byggnader. Uppsatsen har genomförts i samarbete med Vasakronan, ett av Sveriges största fastighetsbolag sett till marknadsandelar, med stort fokus på hållbarhet och att minska elbehovet i deras fastigheter. Syftet med denna studie är att avgöra vilken av tre förvaltningsstrategier: Lapa Effekttoppar, Energi-arbitrage eller Frekvensreglering som genererar de högsta intäkterna för ett solcells- och batterilagringssystem i Vasakronans fastighet Lumi. I den undersökta litteraturen identifierades en avsaknad av en teknoekonomisk modell som utvärderar och jämför dessa tre styrstrategier, där Energi-arbitrage och att Kapa Effekttoppar är etablerade, medan Frekvensreglering är en ny möjlig styrstrategi för kommersiella byggnader. Målsättningen utvärderas genom en teknoekonomisk modell som är utvecklad för de tre olika styrstrategierna i modelleringsprogrammet System Advisor Model (SAM) och Microsoft Excel. Modellen utvärderas för batteristorlekar mellan 0-600 kWh med indata på solinstrålning och elpris för åren 2018-2022. Kvantifierbara utdata från modellen i form av Key Performance Indicators (KPI:er) jämförs mellan kontrollstrategier. Nettonuvärde (NPV) och Intern avkastning (IRR) är de ekonomiska KPI:er med mest fokus på för att avgöra vilken kontrollstrategi som är den mest lönsamma. Resultatet av analysen visade att styrstrategierna att Kapa Effekttoppar och Energi-arbitrage inte når ett positiv NPV eller ett IRR över diskonteringsräntan på 6,5 % för någon av batteristorlekarna över 120 kWh eller valda år. Det främsta skälet till detta resultat är att ett större batteri inte generar tillräckligt mycket intäkter genom dessa styrstrategier för att kompensera för den stora investeringen. Utan något batteri alls, eller med ett litet batteri på 120 kWh, överträffar besparingen i elkostnad från solcellssystemet investeringen och ger därmed ett lönsamt resultat. Frekvensreglerings-strategin genererade ett positivt Nuvärde och IRR över diskonteringsräntan för alla batteristorlekar mellan 240 – 600 kWh för alla år förutom 2021, som var året med lägst solinstrålning. Slutsatsen från resultatet för Frekvensreglering är att intäkterna från ett 120 kWh batteri inte räcker för att täcka investeringskostnaden, men för större batterier överstiger intäkterna investeringskostnaden under hela modellperioden. Den övergripande slutsatsen från är att batterilagring med ett litiumbatteri i kommersiella byggnader i kombination med ett solcellssystem endast är lönsamt om batterikapaciteten budas på FCR-marknaderna. I alla andra fall är ett solcellssystem utan batterilagring att föredra ur en ekonomisk synvinkel. Men nya marknader och potentiella affärsmodeller utvecklas kontinuerligt vilket är något som framtida studier bör undersöka. En mer grundlig analys av marknaderna för frekvensreglering och hur dessa kommer att utvecklas över tiden skulle också krävas för att ytterligare validera resultatet av denna avhandling.

The rapidly evolving Nordic Power System demands enhanced flexibility and robustness in electricity production. The traditional role of hydropower plants in regulating the grid frequency has been challenged by new criteria for dynamic stability, which some units struggle to meet due to their relatively poor dynamic performance. This study addresses this challenge by investigating the potential of integrating optimal energy storage systems with hydropower plants. This study aimed to develop a tool that could streamline the process of converting a traditional hydropower plant into a hybrid unit using an optimal energy storage system. The problem is complex and requires an innovative approach that combines electrical engineering expertise with cutting-edge machine-learning algorithms. A comprehensive hydropower plant model, including governor control and mechanical and hydraulic subsystems, was developed and integrated with an energy storage system model to form a hybrid unit. This model was validated using real power plant data. Three distinct XGBoost Regressor models were trained using data samples generated from the optimized hybrid unit. These models aim to predict power and energy requirements for an optimal energy storage solution, including an estimation of wear and tear reduction. The XGBoost Power Regressor achieved a prediction accuracy of 92 % and the XGBoost Energy Regressor demonstrated a 95 % accuracy. The XGBoost Movement Regressor, indicating wear and tear, boasted an accuracy greater than 99 %. The integration of energy storage systems can significantly mitigate wear and tear on a hydropower plant, with reductions of up to 85 % or more. The results indicate that integrating energy storage systems with hydropower units can substantially enhance the dynamic performance, reduce wear and tear and enable the plants to meet the demanding requirements of providing frequency regulation services in the Nordic Power System. The findings of this study culminate in a robust and user-friendly tool capable of accurately estimating optimal energy storage requirements for any hydropower plant tasked with meeting frequency regulation service demands. / Det nordiska kraftsystemet är under snabb förändring och skiftar alltmera till elproduktion med krav på ökad flexibilitet och tillförlitlighet. Vattenkraftverkens traditionella roll som källa till reglering och stabilisering av nätfrekvensen, utmanas nu av nya krav på dynamisk prestanda och stabilitet. På grund av sina relativt dåliga prestanda har vissa vattenkraftverk svårigheter att uppfylla dessa nya krav. Detta examensarbete behandlar denna utmaning genom att undersöka möjligheterna att integrera optimala energilagringssystem med vattenkraftverk. Syftet med arbetet var att utveckla ett verktyg som skulle kunna effektivisera processen för att omvandla ett traditionellt vattenkraftverk till ett hybridkraftverk med hjälp av ett optimalt energilagringssystem. Detta är ett komplext problem som kräver ett innovativt tillvägagångssätt som kombinerar elkraftteknik med avancerade algoritmer för maskininlärning. En omfattande modell utvecklades för att simulera ett vattenkraftverk med styrsystem, mekaniska och hydrauliska system. Denna kraftverksmodell integrerades med en modell för ett energilagringssystem för att tillsammans bilda en hybridenhet. Modellens validitet verifierades med hjälp av verkliga testdata. Med hjälp av data från simuleringar av den optimerade hybridenheten kunde tre XGBoost-regressionsmodeller skapas för att estimera både effekt och energibehov för ett optimalt energilagringssystem. Utöver detta kunde även en uppskattning av minskning av slitage presenteras. XGBoost Power Regressor uppnådde en träffsäkerhet på 92 % och XGBoost Energy Regressor uppvisade en träffsäkerhet på 95 %. XGBoost Movement Regressor, som indikerar slitage, hade en noggrannhet på högre än 99 %. Integrering med energilagringssystem kan avsevärt minska slitaget på ett vattenkraftverk, med minskningar på upp till 85 % eller mer. Resultaten visar att integrering av energilagringssystem och vattenkraftverk väsentligt kan förbättra den dynamiska prestandan, minska slitage och göra det möjligt för kraftverken att uppfylla kraven för att bidra med frekvensregleringstjänster i det nordiska kraftsystemet. Resultaten av denna studie kulminerar i ett robust och användarvänligt verktyg som kan uppskatta ett optimalt energilagringsystem för ett vattenkraftverk som ska uppfylla kraven för frekvensreglering.

The increasing penetration of variable renewable energy sources in the Nordic Power System is causing frequency quality degradation and has increased the importance of primary frequency control provided by hydropower plants. Hydropower is the world’s largest renewable energy source. Its reliability, controllability and dispatchability along with its large storage volume makes it the most important source for providing frequency regulation in the Nordic Power System. Many hydropower plants offering regulating power have Kaplan turbines which have complex mechanical systems. Furthermore, the frequent and fast mechanical movements of the Kaplan turbines, providing frequency regulation causes the problem of wear and tear in the guide vanes and runner blades of the turbines. Kaplan turbines are suitable for stable operation.   To mitigate this problem, a solution of hybrid hydropower plants combined with  battery energy storage systems is investigated in this thesis, where batteries can take care of fast frequency deviations, allowing for a more stable operation of the turbines. The analysis is based on the FCR-N service offered by hydropower plants, because FCR-N is identified as one of the services that requires very fast changes in the output power of the hydropower plant. Modelling and simulation, data analysis and on-site measurement are adopted as main study methods in this thesis.  The simulation models of a hydropower plant and a hybrid hydropower plant are developed for the analysis. The simulation model of the hydropower plant is validated using data from a typical Swedish hydropower plant. Quantification of wear and tear is the main focus of the study. The performance of the hydropower plant and hybrid hydropower plant are compared in terms of wear and tear of turbines, speed of the response of plants to frequency deviations and number of directional changes during the mechanical movements of the turbine. Finally, it is concluded that, addition of batteries with hydropower plants will reduce wear and tear of the turbines as well as improve the frequency quality in the Nordic Power System.

The main purpose of this thesis was to develop and evaluate Key Performance Indicators (KPIs) and battery usage associated with Lithium-ion Battery Energy Storage Systems (LiBESS) used as Frequency Containment Reserve (FCR). The investigation was based on three of Vattenfall´s LiBESS projects that use the same lithium-ion battery technology but vary in system rating and configuration. It was found that two of the most important KPIs are response time and energy efficiency. The response time describes how fast the system can respond to changes in grid frequency. Additionally, the energy efficiency describes how effectively the system can provide energy storage during service and it can be parametrized into the efficiency of the battery, converter and transformer. The results show that all the considered LiBESS can fulfill the response time requirements of 30 seconds for FCR provision. In the future stricter requirements for the response time in grid stabilization services will most likely be required. Nevertheless, the results showed that a well configured LiBESS can provide response times on the millisecond scale. The energy efficiency evaluation showed that the system energy efficiency decreased from 89% to 85% when the power increased from 50% to 100% of rated power. At 75% of rated power it was found that the converter had the lowest efficiency (92%) based on the analysis of the efficiency of all the system components. It was also found that the power consumed by auxiliary loads was nearly constant for the examined power rates and that it significantly reduced the energy efficiency. Lastly, the battery usage analysis showed that the battery often idles or operates at low power rates if the frequency dead-band of ±10 mHz is applied around the nominal value of 50 Hz. Moreover, the battery usage can be characterized by an average State of Charge of 50% and a maximum Depth of Discharge of 30% during both charge and discharge of the batteries.