The global share of photovoltaic electricity has been expanding at a breakneck pace with CAGR of 27% over the last decade. This resulted in remarkable advancements in photovoltaic technology and design. The performance of the majority of PV modules is determined in laboratories under controlled conditions. Perhaps the climatic conditions on-site are not always stable when the PV module is installed. As a result, modules are subjected to changing environmental stressors such as temperature, relative humidity, UV radiation, and cyclic temperatures. Each of these environmental factors results in a distinct mode and mechanism of degradation, resulting in a loss of performance over time. As a result, determining the rate of degradation of a photovoltaic module is crucial for performance prediction and financial analysis. Primarily, this research identifies a degradation model that is adaptable to various of photovoltaic technologies and designs while maintaining long-term accuracy such model is implemented into PVMD toolbox. We use this model to explore the degradation of photovoltaic modules in a variety of locations around the world, as each site has its own distinct environment. We explore a few well-known climates, namely tropical, steppe, temperate, desert, and cold. According to the study's findings, the lifetime of Cold > Steppe > Desert > Temperate > Tropical and inverse for degradation rates. Financial analysis is critical for determining the value and social impact of a project. Thus, the levelized cost of electricity metric is utilized to determine the economic value of photovoltaic modules under the various climatic conditions indicated previously. To perform this study an LCOE model is incorporated in PVMD toolbox consisting of economic and system variables along with growth rate per year parameter to determine future costs of PV modules. Tropical, steppe, and cold regions are evaluated financially using a variety of economic scenarios and discount rates in order to compare the effect of degradation on LCOE. Tropical zones have the highest LCOE values and the greatest influence on LCOE when the scenario is changed. Cold zones exhibit the greatest range of LCOE values in sensitivity analysis. Finally, as crystalline silicon modules approach their theoretical limit, we discovered the tandem module in quest of more efficient technologies. Due to its broad compatibility with silicon modules, this study focuses on perovskite/silicon tandem modules. Despite this, new research indicates that perovskite/silicon tandem modules are susceptible to degradation. As a result, a model is designed with the objective of assessing the total degradation of tandem modules. We focused on the ability of the perovskite layer degradation to affect the overall degradation of the tandem module by employing normalized power in tandem degradation research. To investigate correlations, the energy production loss owing to degradation in tandem modules is compared to that of an independent crystalline silicon module. Following that, the LCOE of tandem modules is assessed and compared to that of crystalline silicon modules in order to determine the economic impact. According to the study, the perovskite layer in 4T tandem modules degrades at a greater rate than the perovskite layer in 2T tandem modules in order to maintain the same lifetime as independent crystalline silicon modules at the location. When perovskite layers in 2T and 4T degrade at optimal rate, they reach economic equilibrium at 5% and 10% additional module cost over crystalline silicon module.

This thesis deals with the Life Cycle Assessment (LCA) of the manufacture of an EXASUN X-Glass photovoltaic (PV) module to understand its true environmental impact. The methodology and general guidelines for conducting an LCA on a Photovoltaic (PV) system are developed, within the scope of this thesis. The LCA follows a standard framework, proposed by International Organization for Standards (ISO) and International Energy Association (IEA). LCA is a technique used to understand the total environmental impact of all the processes involved in the manufacture of a particular product. It takes into account all production process stages, to check the highest impact involved in production. The results from this LCA study are obtained using the SimaPro software. These results quantify the production impact and aid in understanding where and how to improve the carbon footprint of associated production processes. The environmental impact categories from the ReCiPe 2016 assessment methods were used in this study. All the impact categories such as global warming potential, ozone depletion, photochemical ozone formation, etc. were taken into consideration. This includes all the mid point and end point categories offered by ReCiPe method. Additionally, the cumulative energy demand method was used to calculate the total renewable and non-renewable energies mix, used in the production process. Subsequently, using this information, impact categories such as energy payback time, net energy ratio and GHG emissions rate were obtained. This study is based on the data obtained from Ecoinvent LCA database, accessed through the SimaPro software. This data was then changed according to the inventory update published by IEA PVPS. PV cell data was obtained from literature. However, the remaining materials used in the X-Glass module’s manufacture were noted and their associated data was used to develop an appropriate LCA model. The model incroporated factors including transport, quantity, etc. The functional unit (FU) in this study is 1 m2 of X-Glass production. For better analysis, all the impacts caused by manufacturing 1 X-Glass module were used to compare the results. Results clearly show that the contribution of PV cells production to the environmental impact is high when mono-Si is used by the PV module. The energy requirements and geographical influence of these energy mixes that are involved in the PV cell manufacture are clearly explained. The cells for a typical X-Glass manufacture were modeled with a Chinese energy mix as the cells were produced in China. The energy payback time (EPBT) of X-Glass module installed in the Netherlands are 1.3 years with 73 g CO2/kWh GHG emission rate. The CO2 offset period was found be 3.9 years. These modules were then compared to a similar model developed for an European energy mix. The differences in results, due to the change in energy mix show that he energy payback time for manufacturing the same X-Glass module was found to be 1.3 years, with GHG emission rate of 32 g CO2 eq. The CO2 offset period for these was 1.7 years. The obtained results were cross-validated based on literature, outlining various LCA studies conducted on similar PV technologies. Thus, based on the results, it can be concluded that the environmental performance of X-Glass module was found to improve by 40% when cells from Europe are used as the EPBT of EU based model was found to be 1.3 years with an emission rate of 32 g CO2 eq.

The penetration of renewable energy sources is increasing every year and that poses a lot of challenges in the distribution system. Since they are intermittent in nature, they cannot serve the energy demand as stand alone sources without the help of storage technologies or stable energy sources. As the RES, especially the Photo Voltaic generation units are often connected in the distribution system, their intermittent nature may result in a sudden rise or fall in overall generation. Hence the voltage has to be kept within limits to prevent the failure of the system. Storage technologies plays an important role in keeping the voltage levels under control. From the consumer perspective, the usage of electric vehicles is steadily increasing and most of the vehicles are connected to public charging stations to charge the vehicle battery. Hence there is an interaction between the EVs and distribution grids, which is utilized as a storage option here. In this research, the voltage changes due to the penetration of renewable energy sources in the distribution system is regulated by using the electric vehicles as a storage option. This thesis tries to give an understanding on the strategy applied towards regulation of voltage in the distribution system using electric vehicles. The strategy will serve as a platform for regulating voltage using electric vehicles and it can be adapted to any such distribution network. This strategy takes into account the size and capacity of the battery, battery state of charge and the availability of electric vehicles for a given day. Consensus algorithm serves as the communication platform for information interchange between the electric vehicles connected in the network. Average consensus algorithm estimates the average state of charge of the electric vehicles connected in the distributed network and that has been used to ensure equitable contribution of power among the electric vehicles. Markov model is used to simulate the travel pattern of the electric vehicles. This provides a more realistic and accurate pattern of travel. Markov model is implemented with and without considering time of occurrence and the results shows that the pattern is more accurate when time is considered as a parameter. Considering the generated travel pattern, voltage level of the buses in the network and the output from the consensus algorithm, the voltage has been regulated. The Electric vehicles are used for regulating the voltage when the voltage level is beyond the tolerance limits, and during other times, the electric vehicles charge their batteries. This is executed by controlling the charging and discharging cycles of the electric vehicles. The results are discussed based on different cases and the advantages and limitations of each case is explained in detail. The simulation is not robust since the availability of electric vehicles is uncertain. To make it more robust, monte carlo simulations are used. The probability of occurrence of several events has been estimated using monte carlo simulations and through the results, the availability is predicted to a certain extent. Different cases are implemented and analysed and the results shows that the availability depends on the bus, the location of the bus and the travel behaviour of the electric vehicles. Each case is discussed separately and the degree of uncertainty is reduced at the end.

Renewable energy resources are the most successful, promising, and economical mode of energy generation. Moreover, wind energy is also one of the victorious candidates in generating renewable energy. Currently, High Voltage Direct Current (HVDC) systems are designed to integrate the energy harvested in the Offshore Wind Farms (OWF) to the grid system. The HVDC system uses AC-DC converters, and then with the help of DC cables, the power is transmitted from the offshore to the onshore area, where it again gets converted back to AC. The converters used here are called high voltage converters, where these power electronic components have the ability to sustain high voltages and current. Further, in this thesis, Modular Multilevel Converter (MMC) is used for the conversion of AC to DC and vice versa. The MMC is built using stacked Insulated Gate Bipolar Transistors (IGBTs), where these power electronic components are current sensitive in nature. Thereby, if there is a fault, and if this fault current flows through them, then these components get damaged permanently. Moreover, if the offshore MMC goes down, the OWF system has to undergo a restart. Therefore, the DC Circuit Breaker (DCCB) is used in order to clear the DC fault in an HVDC system. Meanwhile, designing a DCCB for the HVDC system is not easy, one major reason is that there are no natural zero crossings in a DC fault. Also, the fault must be cleared very quickly because DC faults have a very high rate of rising in fault currents. Earlier, the MMC protects itself by using the blocking algorithm. Here, the switching devices turn off temporarily until the fault is cleared. When they are turned off, the fault current flows through the freewheeling diodes, as diodes have the capability to withstand high currents. Indeed, if the MMC is blocked, the purpose of DCCB is not satisfied. Therefore, in this thesis, the performance of the MMC is examined with the presence of DCCBs for a 525 KV system. Besides, there are two types of DCCBs used, in order to evaluate the more suitable breaker for the system. Further, with the initial performance, the fault behavior of MMC is analyzed and illustrated. Consequently, the DC inductance, converter inductance, and arm inductance were modified in order to improve the performance of the system. On the other hand, there are two types of MMC, namely Half Bridge MMC (HBMMC) and Full Bridge MMC (FB MMC). Generally, the HB MMC are called fault feeding converters, whereas the FB MMC are called fault blocking converters. This is due to the topology of the FB MMC, where it completely blocks the fault current by itself. Therefore, the performance of the HB MMC with a DCCB is compared with the FB MMC with blocking protection for a 320 KV network.

The goal of this graduation project was to design a state-of-the-art central farm controller. The designed central farm controller distinguishes itself from prevailing controllers by including a subsystem (the optimisation unit) that contributes to an improvement in power transmission effciency, decrease in maintenance costs and an increase in system reliability/robustness. This thesis describes the design process of the optimisation unit and the verifcation of its feasibility. Originally, the plan was to test the design on a remote terminal unit (RTU), but unfortunately due to the current crisis, this was not possible. Therefore, the implementation and testing were carried out only in MATLAB. Tests were performed using a MATPOWER system model which has been derived from a real wind farm topology. The optimisation unit makes use of a meta-heuristic algorithm to solve an optimal reactive power flow dispatch optimisation problem. In this thesis, the feasibility of this optimisation unit is investigated. Furthermore, it is determined which devices should be controlled and of which the usage is optimised. This makes the treated optimisation problem a multiple objective optimisation. Lastly, the robustness is verified by extending the topology and testing the solutions of the optimisation unit.

A dynamic power system model cannot perfectly predict the behaviour of a physical power system. Model validation tries to answer the question whether the model predicts the behaviour of the physical system sufficiently. This study evaluates if and how a dynamic power system model of the Dutch power grid can be validated using PMU measurements. For this, two approaches to validation of dynamic power system models are compared. They are validation using System Wide Simulation and validation using Hybrid Dynamic Simulation. Both approaches use measurements of disturbances in the power system as a basis for the comparison of the measured and the simulated behaviour of the system. With System Wide Simulation, comparison can be done using measured voltage magnitude, active power flow and reactive power flow. With Hybrid Dynamic Simulation, comparison can be done using measured active power flow and reactive power flow. Identification of model inaccuracies is limited by inaccuracies that are the result of the simulation approach. Inaccuracies are introduced by poor representation of the pre-disturbance operating conditions of the model and poor representation of the disturbance in the simulation. System Wide Simulation is more strongly affected in both cases and is therefore less accurate than Hybrid Dynamic Simulation. System Wide Simulation can be used to gain rough insight into the validity of a model. For accurate model validation and subsequent calibration however, System Wide Simulation falls short, and Hybrid Dynamic Simulation is required. Hybrid Dynamic Simulation however requires PMU measurements to be taken in specifically targeted locations.

This report details the design of an instrumentation system to be used on a skeleton sled. The system will measure several quantities on a skeleton track for the athlete to learn from. This data is stored during the run using several sensors on the sled and processed and visualised immediately afterwards on a mobile device. The project is tailor-made for skeleton athlete Akwasi Frimpong: he requested a way to get some insight in his steering using his knees and shoulders and its results on his achievements. This thesis focuses on the data processing, communication and visualisation of the data and the software integration of all sensors. The communication is done using Wi-Fi and the visualisation is realised with a web page. The whole system is implemented with an ESP32 microprocessor which functions as a Wi-Fi access point and web server for the user interface.

Thin film amorphous silicon based PV systems are a rather new and promising photovoltaic technology with plenty field for technological development and potential for unique applications such as integration within the built environment. Nevertheless, in order for the technology to be established, further technical development is required and economical aspects should be tackled, both in the module and system level. From a commercial point of view, the ultimate goal of a PV system, once installed, is to cost effectively deliver the highest possible energy yield. In reality, the actual energy yield every system delivers is lower than the theoretical rated performance. Location and environmental factors, such as heating of the modules, irradiance, dirt accumulation, angle of incidence, Etc. affect the performance of the systems. Understanding the behavior of this new technology under real operation conditions can provide reliable insights for its further development, help better estimate the energy yield amorphous silicon systems can provide and perhaps, help the technology differentiate itself from the better commercially established crystalline silicone based one. This work investigates the effect of most of the location and environmental factors affecting the performance and the ultimate energy yield of a photovoltaic system comprised of flexible and lightweight amorphous silicon modules. The factors responsible for this performance decrease are individually investigated, and its effects are quantified by addressing in which manner they decrease the performance ratio of the system. A performance monitoring system was built expressly for this purpose, which data acquisition plan was based on the IEC 61724 -1 (2017) standard. A computer model with the capability of simulate the performance of the system after correcting for the evaluated mechanism was also constructed as part of this work using Matlab. This model was based on an experimental characterization of the modules, using live performance data as an input to give a performance ratio, and an assessment of the effect of loss factors. The results allocated the source of the system losses up to a 78% of the total power decrease, considering the rest of the unexplained observed losses as effects of the active material degradation. The performance ratio of the system was reported on a daily bases for the full evaluated period, concluding a strong correlation between a high irradiance and a high performance ratio. The common literature claim of a better behavior of thin film a-Si:H technologies compared to c-Si was verified under this study, and the effects of system operation under low irradiance conditions was identified as the dominant loss mechanism.

Bifacial PV modules and floating PV systems are two fast-growing technologies in the photovoltaic sector. Bifacial PV modules make use of the irradiance that is incident on both faces of the module, leading to higher energy yields. Several techniques are used in order to enhance this bifaciality effect, one of them being the use of reflectors. On the other hand, floating PV systems offer multiple advantages with respect to conventional in-land PV systems such as land saving (a big issue for vastly populated areas), higher efficiencies and lower water evaporation rates. The advantages that these technologies provide as well as the big market opportunities that they offer, make them very attractive for new investors.
The uncertainties related to floating PV and bifacial modules are higher compared to conventional power plants. In order to minimize these uncertainties, accurate modelling tools are essential. The most widespread software suites such as PVSyst or System Advisor Model, offer limited flexibility to the user. Therefore, the user is not able to adapt the simulation to very specific cases such as the addition of reflectors to the system. Additionally, they perform simulations only at module level, without performing an analysis at cell level.
Aiming to fill this gap, a Toolbox that is able to model and simulate the annual energy yield of a PV module, considering the physical effects from cell to module level, was developed by the Pho- tovoltaic Devices and Materials group of TU Delft. This Toolbox allows the integration of the work done by different members of the group and it offers great simulation flexibility to the user. The goal of this thesis is to continue the work carried out by former PVMD members and create a new and improved version of the Toolbox.
During this thesis project, several improvements were carried out in the optical models of the Tool- box. Multiple mistakes were discovered and corrected, and new features that improve the Toolbox were added, making it suitable for its use in real projects. Besides that, the thermal part has been improved by adding the option to utilize a thermal model developed in COMSOL Multiphysics that allows to study the temperature distribution in a bifacial PV module. Finally, the Toolbox was ap- plied to the INNOZOWA project, where several simulations were carried out to find the ideal design of a floating PV plant that consists of bifacial PV modules and reflectors. It was found that the chosen design for this type of system can provide a bifacial gain of 18%.