A transition phase is taking place in Sweden, where the goal is to become a completely climate neutral country by 2045. The transport sector currently accounts for a third of fossil emissions in Sweden, while the transport sector also has the greatest potential to become fossil-free through, forexample, a comprehensive electrification. Bottlenecks in the grid is a challenge that Sweden faces where the existing ability to send powerthrough the country is already highly utilized. Battery storage can partly be the solution to this problem and also support the future needs that a further electrification of the transport sector may cause. Battery storage can, however, be both expensive and require a lot of space. To avoid this, the mobile battery storage that is available in electric cars can be used to convey power to the grid based on need. The technology that performs this bidirectional charging is called V2G (vehicle-to-grid) and has enormous theoretical potential. The number of electric cars in Sweden has increased by 82% duringthe year 2020, which provides good conditions for continuing to investigate the potential for V2G. Previous studies have shown challenges with the technology. The main issues pointed out have been profitability, winter conditions and battery wear, all of which are taken into account in this study. As in all of Sweden, Halmstad needs to plan for its electrification of the transport sector and load consequences on the grid. This study carries out a combined qualitative and quantitative case study that examines how a future electric car fleet can affect Halmstad's local grid. With data from HEM Nät from a winter week that will correspond to extreme conditions for the grid, a model has been developed in Excel to test different proposed scenarios. What is analyzed is how V2G can work in practice depending on where and when charging takes place, and whether power regulation can be profitable for both private individuals and network operator. Results show that some form of power regulation will be needed in the future to deal with the consequences of uncontrolled electric car charging with an ever-larger electric car fleet, and that V2G may be an option. Despite the winter climate and consideration for battery wear, a significant power peak reduction can be achieved if sufficient participation is attained and a good control strategy is found. Financial analysis shows a negative outcome for private individuals who use V2G. The utility services that is provided can on the other hand reduce costs for the network operator through load balancing incentives and reduced subscriptions to overlying networks. This in turn can enable an interest in network operators to introduce local incentives for private individuals' involvement.

Electric vehicles, which have both energy storage capability and mobility capability, can provide a new solution for electricity sharing between different building communities (i.e., a group of buildings connected with a microgrid). This comes to the community-to-community (C2C) energy sharing network. The C2C energy sharing networks have the potential to not only minimize the effects of electric vehicle integration into the energy grid, but also improve the electricity grid efficiency as a whole. In this thesis, a coordinated smart charging method of electric vehicles (EVs) is proposed for the C2C model. The proposed method considers the power regulation needs in both the present parking community and the next destination community. Then, based on the needs of both communities, the control method will decide the optimal amount of electricity that can be delivered by EV, so that the energy performances in both communities can be the best. The developed coordinated control has been compared with a base case (without any smart charging) and an uncoordinated control case under two control strategies: minimizing the peak energy exchanges with the grid and maximizing the renewable self-utilization. The genetic algorithm tools in MATLAB software are used for the optimization of the model. Meanwhile, to test the robustness of this C2C model, different combinations of building communities have been studied, namely residential-workplace, residential-university, and residential-workshop communities. The case study reveals that the C2C model is effective in improving energy performance under both control strategies. Peak reduction control strategies work most effectively for smaller systems with lower electricity demand and production. With C2C energy sharing, the annual mean peak reduction ranged from 39 % at the smallest community and 20 % at the largest community. Self-consumption maximization strategies work best for systems with a larger surplus of electricity production. With C2C energy sharing, the annual self-consumption increase ranged from 50 % at the community with the largest production surplus, to 7 % at the community with the smallest production surplus. The residential-workshop community studied in this thesis benefited the most from C2C charging control due to its production surplus and the relatively low electricity demands of the communities.

Battery energy storage systems are a key factor for enabling a continuous increase of the fraction of photovoltaics in the Swedish electricity grid. One big challenge is to utilise all potential services of such a storage system. The aim of this study was to improve the supervisory controller for an existing battery storage and photovoltaic solution marketed by the Swedish company Ferroamp AB. This has been done by developing a combined peak reduction and time-of-use bill management algorithm, together with a simulation and evaluation software for optimisation of algorithm parameters. The algorithms and tools were evaluated using an installation made by Ferroamp AB and Vattenfall Eldistribution AB as a case study. Sensitivity analyses has been performed on economic parameters and length of the algorithm training data set. Improvement of economic profit, in this case study, were 300 % compared to the currently used algorithm and 32 % compared to a conventional threshold peak reduction algorithm. Despite this improvement, the battery energy storage system is shown to be non-profitable, with the economic profit only covering 36 % of the investment costs, not taking interest rate into account. Like in many other studies, power storage was found more profitable than energy storage. An increase of the grid power tariff and the grid energy fee of 30 % to 40 % is found to make the system viable. One interesting finding is that by using the proposed optimal algorithm, 55 % of the cycle life of the battery storage is still accessible for other services when considering 10 years of economic depreciation time for the system.

Dans cette thèse, nous nous intéressons à la résolution parallèle de grands systèmes linéaires creux. Nous nous focalisons plus particulièrement sur les solveurs linéaires creux hybrides directs itératifs tels que HIPS, MaPHyS, PDSLIN ou ShyLU, qui sont basés sur une décomposition de domaine et une approche « complément de Schur ». Bien que ces solveurs soient moins coûteux en temps et en mémoire que leurs homologues directs, ils ne sont néanmoins pas exempts de surcoûts. Dans une première partie, nous présentons les différentes méthodes de réduction de la consommation mémoire déjà existantes et en proposons une nouvelle qui n’impacte pas la robustesse numérique du précondionneur construit. Cette technique se base sur une atténuation du pic mémoire par un ordonnancement spécifique des tâches de calcul, d’allocation et de désallocation des blocs, notamment ceux se trouvant dans les parties « couplage » des domaines.Dans une seconde partie, nous nous intéressons à la question de l’équilibrage de la charge que pose la décomposition de domaine pour le calcul parallèle. Ce problème revient à partitionner le graphe d’adjacence de la matrice en autant de parties que de domaines désirés. Nous mettons en évidence le fait que pour avoir un équilibrage correct des temps de calcul lors des phases les plus coûteuses d’un solveur hybride tel que MaPHyS, il faut à la fois équilibrer les domaines en termes de nombre de noeuds et de taille d’interface locale. Jusqu’à aujourd’hui, les partitionneurs de graphes tels que Scotch et MeTiS ne s’intéressaient toutefois qu’au premier critère (la taille des domaines) dans le contexte de la renumérotation des matrices creuses. Nous proposons plusieurs variantes des algorithmes existants afin de prendre également en compte l’équilibrage des interfaces locales. Toutes nos modifications sont implémentées dans le partitionneur Scotch, et nous présentons des résultats sur de grands cas de tests industriels. / In this thesis, we focus on the parallel solving of large sparse linear systems. Our main interestis on direct-iterative hybrid solvers such as HIPS, MaPHyS, PDSLIN or ShyLU, whichrely on domain decomposition and Schur complement approaches. Althrough these solvers arenot as time and space consuming as direct methods, they still suffer from serious overheads. Ina first part, we thus present the existing techniques for reducing the memory consumption, andwe present a new method which does not impact the numerical robustness of the preconditioner.This technique reduces the memory peak by doing a special scheduling of computation, allocation,and freeing tasks in particular in the Schur coupling blocks of the matrix. In a second part,we focus on the load balancing of the domain decomposition in a parallel context. This problemconsists in partitioning the adjacency graph of the matrix in as many domains as desired. Wepoint out that a good load balancing for the most expensive steps of an hybrid solver such asMaPHyS relies on the balancing of both interior nodes and interface nodes of the domains.Through, until now, graph partitioners such as MeTiS or Scotch used to optimize only thefirst criteria (i.e., the balancing of interior nodes) in the context of sparse matrix ordering. Wepropose different variations of the existing algorithms to improve the balancing of interface nodesand interior nodes simultaneously. All our changes are implemented in the Scotch partitioner.We present our results on large collection of matrices coming from real industrial cases.