The master thesis will concern design of hybrid photovoltaic system for a family house with accumulation of electric energy. Three alternatives of power consumption will be made. For each alternative multiple simulations will be carried out with different amount of photovoltaic panels and batteries. The most appropriate alternative and size of photovoltaic system will be chosen based on those simulations. Finally the choice of alternative will be described.

This diploma thesis provides basic information about the thermal pumps and photovoltaic systems. The heat pumps are described regarding their construction and sources of low potential heat. The description of photovoltaic systems explains the main principle of transformation the solar light to electric power and possible cooperation of this systems with the heat pumps. In practical part of the thesis there is a description of the assessed building and calculation of its thermal losses. It is followed by three designs of the thermal pump systems which are supposed to replace the original way of heating - the electric boiler. One of the designs is supported by photovoltaic system. The final part of the thesis provides the calculation of required heat sources for house and water heating, the operation expanses of all three heat pump designs, prices of electric energy and economical reliability of described solutions.

This diploma thesis deals with available technologies when designing autonomous house. Theoretical part of this thesis shows and presents options for gaining energy from renewable resources and options for energy storage. The practical part of the thesis is focused on designing partly autonomous system for model house using renewable resources. In the thesis are three options of gainig energy for the house, from which is chosen the best option.

The diploma thesis studies the elaboration of project about optimization of solar system for all- season service. The aim of the work was to make a proportioning, choosing the right parts and finding the economic and ecologic balance.

This master's thesis deals with a design of backup power system, which will be using renewable energy sources, particularly solar energy. The first part describes the potential of solar power plant in the Czech Republic. The next parts describe types of photovoltaic systems, their components, design of photovoltaic systems and possibilities of controlling power consumption. Due to the fluctuating supply from renewable energy sources the thesis also deals with possibilities of predicting of the production electricity from these sources. In the practical part the design of backup power system for the house is done, which will also be working in summer as an optimizer for own consumption. A financial evaluation was done for this proposal.

The increased usage of solar panels worldwide highlights the importance of being able to detect faults in systems that use these panels. In this project, the historical power output (kWh) from solar panels combined with meteorological data was used to train a machine learning model to predict the expected power output of a given solar panel system. Using the expected power output, a comparison was made between the expected and the actual power output to analyze if the system was exposed to a fault. The result was that when applying the explained method an expected output could be created which closely resembled the actual output of a given solar panel system with some over- and undershooting. Consequentially, when simulating a fault (50% decrease of the power output), it was possible for the system to detect all faults if analyzed over a two-week period. These results show that it is possible to model the predicted output of a solar panel system with a machine learning model (using meteorological data) and use it to evaluate if the system is producing as much power as it should be. Improvements can be made to the system where adding additional meteorological data, increasing the precision of the meteorological data and training the machine learning model on more data are some of the options. / Med en ökande användning av solpaneler runt om i världen ökar även betydelsen av att kunna upptäcka driftstörningar i panelerna. Genom att utnyttja den historiska uteffekten (kWh) från solpaneler samt meteorologisk data används maskininlärningsmodeller för att förutspå den förväntade uteffekten för ett givet solpanelssystem. Den förväntade uteffekten används sedan i en jämförelse med den faktiska uteffekten för att upptäcka om en driftstörning har uppstått i systemet. Resultatet av att använda den här metoden är att en förväntad uteffekt som efterliknar den faktiska uteffekten modelleras. Följaktligen, när ett fel simuleras (50% minskning av uteffekt), så är det möjligt för systemet att hitta alla introducerade fel vid analys över ett tidsspann på två veckor. Dessa resultat visar att det är möjligt att modellera en förväntad uteffekt av ett solpanelssystem med en maskininlärningsmodell och att använda den för att utvärdera om systemet producerar så mycket uteffekt som det bör göra. Systemet kan förbättras på några vis där tilläggandet av fler meteorologiska parametrar, öka precision av den meteorologiska datan och träna maskininlärningsmodellen på mer data är några möjligheter.

Due to the significant increase of ’renewable energy’ to the total energy the highly fluctuating energy supply, which is due to the constitutional conditions of production among others from photovoltaic systems, becomes a growing problem. As a result, many engineering firms and companies dedicate themselves to the so-called power prediction by which it should be possible to integrate a fast changing energy supply into a necessarily continuous energy supply. The Ingenieurbüro für Last- und Energiemanagement (LEM-Software) used a neural network for this prediction of performance of photovoltaic systems. This networks learn from past knowledge a mathematical patern that can be used for forecasting. Based on this, this article deals with a new parameter for the network which characterizes the probability of broken cloud effects. The global radiation forecast in the surrounding area is analyzed for this index. It turns out that this inhomogeneous index is a better indicator than the standard deviation. A first application shows a slight improvement in the forecast result. However, the time limit for application of the neural network is too short for a final evaluation. / Aufgrund des steigenden Anteils ’erneuerbarer Energie’ an der Energieversorgung wird die stark schwankende Energiezufuhr, bedingt durch die naturgegebenen Produktionsbedingungen unter anderem von Photovoltaik-Anlagen, zu einem immer größeren Problem. Infolge dessen widmen sich viele Ingenieurbüros und Firmen sogenannten Leistungsprognosen, mit deren Hilfe es gelingen soll, eine schnell wechselnde Energiebereitstellung in eine notwendigerweise kontinuierliche Energieversorgung zu streichen. Das Ingenieurbüro für Last- und Energiemanagement (LEM-Software) verwendet für diese Vorhersagen der Leistung von Photovoltaik-Anlagen ein künstliches neuronales Netzwerk (KNN). Dieses erlernt aus vergangenen Verhältnissen ein mathematisches Muster, welches für die Prognose angewendet werden kann. Aufbauend darauf befasst sich dieser Artikel mit einem Inhomogenitätsindex f¨ur das Netzwerk, welcher die Wahrscheinlichkeit für das Auftreten schnell wechselnder Bewölkung und von Broken-Cloud-Effekten charakterisiert. Für den Index wird die Globalstrahlungsvorhersage in der näheren Umgebung analysiert. Dabei stellt sich heraus, dass dieser Inhomogenitätsindex ein besserer Indikator als die Standardabweichung ist. Eine erste Anwendung zeigt eine leichte Verbesserung des Prognoseergebnisses, allerdings ist der Zeitraum für die Anwendung des KNN zu kurz für eine abschließende Bewertung.

This study evaluates energy performance and economic analysis of variable refrigerant flow (VRF) systems in U.S. climate locations using widelyepted whole building energy modeling software, EnergyPlus. VRF systems are known for their high energy performance and thus can improve energy efficiency in buildings. To evaluate the energy performance of a VRF system, energy simulation modeling and calibration of a VRF heat pump (HP) type system is performed using the EnergyPlus program based on measured data collected from an experimental facility at Oak Ridge National Laboratory (ORNL). In the calibration procedures, the energy simulation model is calibrated, according to the ASHRAE Guideline 14-2014, under cooling and heating seasons. After a proper calibration of the simulation model, the VRF HP system is placed in U.S. climate locations to evaluate the performance variations in different weather conditions. An office prototype building model, developed by the U.S. Department of Energy (DOE), is used with the VRF HP system in this study. This study also considers net-zero energy building (NZEB) design of VRF systems with a distributed photovoltaic (PV) system. The NZEB concept has been considered as one of the remedies to reduce electric energy usages and achieve high energy efficiency in buildings. Both the VRF HP and VRF heat recovery (HR) system types are considered in the NZEB design, and a solar PV system is utilized to enable NZEB balances in U.S. climate locations by assuming that net-metering available within the electrical grid-level. In addition, this study conducts life cycle cost analysis (LCCA) of NZEBs with VRF HP and HR systems. LCCA provides present values at a given study period, discounted payback period, and net-savings between VRF HP and HR systems in U.S. climate locations. Preliminary results indicate that the simulated VRF HP system can reasonably predict the energy performance of the actual VRF HP system and reduce between 15-45% for HVAC site energy uses when compared to a VAV system in U.S. climate locations. The VRF HR system can be used to lower building energy demand and thus achieve NZEB performance effectively in some hot and mild U.S. climate locations.

M. Tech. (Electronic Engineering, Faculty of Engineering and Technology), Vaal University of Technology| / The lack of electricity in off-grid rural communities in South Africa (SA) and Africa as well as the budget constraints on these communities restrict these communities from connecting to any online resources (internet and e-books) for educational purposes, thus creating a major contributor to the global digital divide. Renewable energy sources such as solar energy, wind energy and biomass were presented as potential alternatives to grid-connected electricity for remote rural locations. Solar energy was identified as the amply available alternative energy resource in SA. Solar radiation was converted by photovoltaic technology to electricity. National power grid isolation (off-grid) was achieved by using a stand-alone photovoltaic system. Photovoltaic technology classification, material categorisation and system sizing for an e-learning centre was presented. Practical set-ups were utilised to determine the most cost-effective equipment mix of power utilization, power management/storage and ICT equipment to build a pilot e-learning centre. It was established that one photovoltaic panel can be employed to fully recharge a battery of a pilot e-learning centre with an operational period of 7 hours using the available sunlight hours. Owing to the susceptibility of the Vaal Triangle region to thunderstorms causing overcast conditions for days, a ratio of back-up battery capacity (Ah) to number of days (seven hours per day) without sunlight was determined. An algorithm was also derived for sizing the pilot e-learning centre for full scale implementation. Future research recommendations based on this study for a reduced system costs of an off-grid e-learning for rural communities powered by a renewable energy resource were presented. This will increase access to basic education in SA and reduce the global digital divide.

Au Vietnam, plus d’un demi-million d’habitations n’ont pas d’accès à l’électricité. Elles se situent principalement dans des régions montagneuses ou sur des îles. Cependant, c’est un pays qui possède un grand potentiel en énergies renouvelables. Dans ce contexte, l’alimentation en électricité pour les sites isolés est une solution prometteuse en termes économique et environnemental. L’énergie solaire est la plus adaptée à l’alimentation en électricité des villages en raison d’un ensoleillement important et d’une maintenance relativement facile. Dans les systèmes de conversion d’énergie utilisant des sources d’énergies renouvelables, on utilise généralement des convertisseurs statiques simples. En effet, si l’on prend un système photovoltaïque, la poursuite du point de puissance maximale (MPPT) se fait à l’aide d’un convertisseur « boost » ou « buck-boost ». Ainsi, en cas de défaillance, le système est mis hors service. L’objectif de cette thèse est d’apporter des améliorations au niveau d’un système photovoltaïque autonome pouvant être utilisé dans un site isolé. Ainsi nous avons développé un algorithme de recherche du point de puissance maximale (MPPT) utilisant des convertisseurs DC-DC à trois niveaux (CBTN) permettant d’extraire le maximum de puissance d’un générateur photovoltaïque quelles que soient les variations climatiques (température, ensoleillement) ou de la charge. L’architecture à base de panneaux solaires associé à un système de stockage a nécessité la mise en place d’un superviseur flou afin de maîtriser la gestion des flux. Enfin, nous avons proposé une méthode de détection de défauts afin de gérer efficacement les cas de défaillance d’un élément du convertisseur multiniveaux. En effet, en cas de défauts, on doit pouvoir passer en mode dégradé pour pouvoir assurer un service proche du comportement nominal ou au moins minimal en attendant une maintenance corrective. / In Vietnam, more than half a million people do not have access to electricity. They are mainly in mountainous regions or on islands. However, this country has great potential for renewable energy. In this context, these sources of energy can be regarded as promising solutions both economically and environmentally for supplying electrical power. Solar energy is the most suitable to supply villages with electricity because of the plentiful solar radiation and relatively easy maintenance of the structures. In energy conversion systems using renewable energy sources, simple static converters are generally used. Indeed, if we explore a photovoltaic system, the maximum power point tracking (MPPT) is done using a boost or buck-boost converter. Thus, in case of failure, the system is simply switched off. The aim of this thesis is to make improvements to an autonomous photovoltaic system that can be used in an isolated site. Therefore, an MPPT algorithm using three-level DC-DC converters is developed to extract the maximum power of a photovoltaic generator, whatever the climatic variations (temperature, sunlight) or charge. The system’s architecture is based on solar panels associated with a storage system, which required the development of a fuzzy supervisor to control the flow management. Finally, we propose a fault detection method to efficiently manage the failure of a multi-level converter element. Indeed, if there is a fault, we must go into a degraded mode to provide a service close to normal or at least minimal functioning, pending maintenance.