While the renewable energy sources are rising more and more throughout the years, and the effects of climate change starting to become more visible in our everyday life, some actions need to be taken. Combining different renewable energy sources gives the opportunity to build systems that can help limit these effects and provide energy in different types. The main goal of this project is to model a system that can offer electricity and hydrogen to a steel industry, in this case Tata Steel, and provide an insight for their next steps towards sustainability not only to cover their electricity needs but also investigate alternative steel making routes for their primary processes which are the iron making and further the steel making with hydrogen injection. The models developed for this project aim to answer the following research question: ”What is the feasibility to cover with a solar - wind - hydrogen system, the loads of the different steel making routes of Tata Steel ?” Different models were build through the software TRNSYS for the scenarios investigated that are linked with different combinations of components and loads so more sustainable steel making routes can be followed in the future. The steel making routes refer to the iron making process and the electrification of the industry with green electricity. For the year of 2030 a fuel mix of 70% natural gas and 30% of hydrogen and for the year of 2050, 100% of hydrogen will be used for the iron making process while for both of them 100% of the electricity loads were investigated to be covered by renewable energy sources. The scenarios were divided in a combining structure of local - non local generation and the loads corresponding to the different steel making routes. The local scenario refers to generation in the Netherlands while the non local scenario refers to the generation of hydrogen in the Arabic Peninsula and the generation of the electricity loads of the processes in the Netherlands. The components that were used were wind turbines, solar panels, batteries and electrolyzers. Each system was optimized with the help of GenOpt, an add - on of TRNSYS, to which was set to minimize the levelized cost of electricity, considering also the load coverage both for the hydrogen and electricity loads not to deviate more than 1% from the full coverage. Further than the optimization, a sensitivity analysis with the Sobol method through the programming environment of python and more specifically the SALib library and the parametric analysis of TRNSYS was conducted for all the different scenarios investigated to give an insight how the amount of the different components affects the levelized cost of electricity. Combining different metrics that were calculated as the Levelised cost of electricity (LCOE), the Self Sufficiency Ratio (SSR), the SSR of hydrogen, the total cost, the avoided emissions per MWh, the area ratio and the avoided emissions per area, a final comparison was done through the different scenarios. The most attractive choices both in a feasibility and economical perspective were the scenarios for local generation of hydrogen and electricity for the projected steel making routes of the years of 2030 and 2050. For the local generation scenario of 2030 an LCOE of 0.383 AC/kWh with an electricity coverage of 99.007 % and a hydrogen load coverage of 99.135% were resulted. On the other hand, for the scenario of the local generation of 2050, an LCOE of 0.421 AC/kWh, with an electricity load coverage of 99.875% and a hydrogen load coverage of 99.207% were calculated. The scenario that was the most attractive throughout the different metrics was the one for the local generation of 2050 while the next to come was the one for the local generation of 2030. Given the aforementioned study and its respectful results, it is a first step to evaluate the impact that such a system can have not only in the emissions reduction of such an intensively emitting industry but also to bring into perspective all the different aspects needed to be considered to realize such a project and evaluate them in more depth in the future.

The world is in need for an energy transition from the usage of fossil fuels to renewable energy source to reduce the harmful effects climate and health. Focusing on accelerating the research and development of photovoltaic technologies is an important step towards energy transition. The thin-film silicon solar modules has shown a great potential and a high market prospect. It is flexible, light-weight and also has a low manufacturing and installation cost. It is used for solar pumps and building integration. However, the highest initial conversion efficiency for a micromorph is 14.8%. Thus, we focus on further optimizing the multi-junction solar cells to reach higher efficiency by focusing on the improving the p-layer, texturing and silicon oxide intermediate reflector(SOIR). Further, standardized and error-free external quantum efficiency (EQE) measurement was done to obtain reliable Jsc values.

This work introduces deposition of single, micromorph and triple-junction thin-film solar cells on glass and wafers. The experiments were designed to study the effects of varying the material, thickness of the p-layer and n-SiOx reflective layer and textures of the solar cells. The Jsc from the EQE, Voc, FF and efficiency were the main parameters used to analyze the results.

The Jsc from the EQE measurement was not reliable as the EQE had an artifact while measuring micromorph devices. It was solved by increasing the bias light intensity and decreasing probe light intensity. The forward voltage bias was also found to be important for a cell with low shunt resistance and to measure the target cell in short circuit condition. The standard bias lights to be used are 1-3 to bias the top cell and 7-8 to bias the bottom cell combined with 50% reduction in the probe light intensity.

After correcting the EQE, the optimization of the top cell p-layer was performed on a-Si single junction solar cells on Asahi substrates. It was found that both the contact layer and the window layer of the p-layer have a high influence on the performance of the solar cell. A maximum conversion efficiency of 10% and Voc of 910mV were achieved for the a-Si single junction solar cells. Further, it was also found that there is more than one way to reach high efficiency with an optimal p layer by changing the material composition and thickness of both the contact and window layer. Additionally, p-SiOx contact layer resulted in good performing solar cells for larger range of F_B2H6 of the window layer than p-nc-Si contact layer and the insertion of the buffer layer in the i/p interface showed no significant improvement in the performance of the solar cell.

Furthermore, comparison of the performance of textures of different feature sizes was done on solar cells in p-i-n and n-i-p architecture. It was found that the best optoelectronic performance was obtained from asahi and smooth pyramidal textured solar cell in p-i-n and n-i-p configuration respectively in both double and triple junction solar cell. Further, the honeycomb textured substrate with R_rms around 270nm is not suitable to deposit on double and triple junction solar cell as it resulted in shunted cells.

Finally, the SOIR can be used in a-Si/nc-Si/nc-Si triple junction solar cell to divide current within middle and bottom junctions. The reflection of light in the infrared region increases with the thickness of the SOIR layer. A thick 90nm n-SiOx layer with F_B2H6=3.2sccm gives the best electrical performance and current distribution property for a micromorph.

One of the goals outlined during the Paris Agreement in 2015 aimed at 'holding the increase in global average temperature to well below 2C above pre-industrial levels and pursuing efforts to limit the temperature increase to 1.5C'. In conjunction with this, the Klimaatakkord of the Netherlands aims to 'reduce greenhouse gas emissions in the Netherlands by 49% compared to 1990 levels' To achieve this goal, a rapid decarbonisation of our economy and energy system is needed. Currently, residential usage accounts 20.4% of Dutch energy consumption.. To reach these targets, the integration of renewable energy sources in Dutch households will be a needed.

Solar energy is already one of the most affordable renewable energy sources available and is currently being integrated into newly built households across the Netherlands. However, as the renewable capacity of the Netherlands expands, so will the need for energy storage to meet the mismatch between renewable generation and demand. A battery bank is usually adopted to supply this mismatch on a daily basis and the production and consumption of hydrogen the chosen technology for a seasonal one. Thus, future households and neighbourhoods in the Netherlands must incorporate both in order to maximise self sufficiency from the grid. The high costs of these components make it unsuitable for implementation in a single household, but scaling up to provide for an entire neighbourhood is a more feasible approach. This results in a so called grid-connected hybrid PV-Battery-Electrolyser-FC energy system.

This final thesis project models and optimises a grid-connected hybrid PV-Battery-Electrolyser-FC energy system to asses its feasibility, both economically and technologically, for utilisation on a neighbourhood in the Netherlands. The simulation model of the hybrid energy system is designed TRNSYS. The model is optimised to minimise the levelised cost of eletricity (LCOE) and to maintain a self-sufficiency ratio of 1\% for the hybrid energy system in TRNOPT. Several scenarios are optimised based on the overall system layout and cases dependant on the electrical, heat and mobility demand. The particle swarm optimisation (PSO) and Hooke-Jeeves optimisation algorithms are used for the optimisation process in GenOpt. In addition, a literature study on the learning curves of different components in the hybrid energy system was performed to predict their costs in 2030. The results of this were used to optimise the system as if it were built in 2030.

The simulated hybrid PV-Battery-Electrolyser-FC energy systems are technically feasible for most scenarios and load profiles for a Dutch neighbourhood. The one exception to this is heat load demand with de-centralised PV generation, which saw an energy deficit at the end of the year. The lowest LCOE of 0.749 €/kWh was found for the centralised scenario implementing smart load management in the load demand. It is found that de-centralising the PV-system to the roofs of houses and the battery storage system each increases of the LCOE of the system due to larger installations costs and a different battery technology. The preliminary results of the future scenarios suggest the results will follow the same trends as was seen in 2020. The LCOE reduces by 21% - 28% compared to the LCOE of 2020. However, more research is needed on this topic to draw conclusive results.

The maximal conversion efficiency on a single pn-junction solar cell is calculated to be 33.7% at a band gap energy of 1.34 eV. The main reasons for this maximum are the non-absorption of low energy photons and the thermalization of high energy photons. To counter this, multijunction devices can be constructed to make optimal use of the spectral range of the sunlight reaching the earth’s surface. While the introduction of thin film multi-junctions based on
silicon has managed to overcome the problem of spectral mismatching, photons with an energy below 1.12 eV are still not able to be collected. Where usually a high current density is traded off against a high open circuit voltage in multi-junction devices, the addition of a low band gap germanium-based bottom cell (Eg = 0.67 eV) could improve open circuit voltage without limiting the current density in the device. The photon flux in the range of 0.67-1.12 eV is high enough such that a current density of 15.9 mA/cm2 can be produced when they can be effectively collected. This current density is so high that it will never limit the output current of a multi-junction device, making germanium a very attractive material for studying and integration in structures where the low energy photons are not yet utilised.

This thesis is focused around the growth and characterization of hydrogenated germanium films for future use as active absorbers in multi-junction deivces. The influence of a set of deposition parameters on the morphological, optical and electrical properties of the films is studied with the aim of finding a deposition regime where device quality germanium films can be produced within the CASCADE reactor in the EKL clean room. The thin films are all be deposited using the RF-PECVD technique. Uniformly deposited Ge:H films under a stable
plasma can be deposited in CASCADE in the range of 1-5 mbar pressure and 5-30 W RF power at a fixed electrode distance of 20 mm. The films start to crystallize and form nc-Ge:H at a pressure of around 3-4 mbar and an RF power of 15-25 W when the germane precursor gas is diluted with hydrogen in a ratio of 1:400. A strong correlation between the refractive index and the presence of post deposition oxidation is investigated. Films with a low refractive index are characterized as having a low material denisty, making it easy for ambient water
molecules to diffuse into the lattice and react with germanium dangling bonds present there. By studying the effect and extent of the post-deposition oxidation on other film properties it was found that the activation energy of the films decreases to values as low as 50 meV. Despite this, due to a significant decrease of the σ0 by 1-5 orders of magnitude, the dark conductivity is found to decrease by 1-3 orders of magnitude. With a high photo/dark conductivity of 5-6 as a result. These results have led to the belief that the formation of Ge-O bonds in the films decreases the amount of defective states in the film, but that the defect states are moved up to an energy level closer to the conduction band. The presence of Ge-O bonds also inhibits the development of low band gap absorbers as seen by the low E04 optical band gap. In the denser films without oxygen contamination, the lowest E04 that has been reported is 1.2 eV along with a Tauc gap of 0.93 eV and a photo/dark conductivity ratio of 3.4.

Single junction solar cells have a theoretical efficiency limit of 33.1% due to spectral mismatch. To overcome this, multi-junction devices are generally fabricated with two or more junctions, so as to achieve better energy conversion efficiency by optimum spectral utilization. The c-Si bottom cell of a thin-film Si triple-/quadruple-junction device does not utilize the low energy photons (below 1.1 eV). The photons in the range of 0.7-1.1 eV, have an available current density of 15.9 mA/cm^2. A fraction of this current density would be large enough to not limit the output current of these thin-film Si-based multi-junction devices. Ge, belonging to the same group IV as Si and being heavier than it, forms weaker covalent bonds. Hence, it has a lower bandgap energy, making it the preferred choice of material. In this work, a low bandgap material such as a Ge-based absorber layer is fabricated that can be used in the bottom cell of a thin-film Si-based quadruple-junction device. This thesis will focus on the influence of a set of deposition parameters on the various properties of the Ge:H films. This will result in a set of Ge:H films from which a specific few are used as absorber layers to analyze the performance of a single-junction cell. The fabrication of the Ge:H films is carried out on a CASCADE setup which is based on the RF-PECVD technique. A processing range is identified to be in the range of 1-5 mbar pressure with RF powers ranging between 5-25 W for a fixed electrode distance of 20mm. nc-Ge:H are processed in the range of 20-25 W for pressures of 2 mbar and higher at a high dilution of 400. A strong correlation is found between the refractive index of the films and the presence of GeOx. The films with low refractive index possibly indicate a porous network with high void density show substantial oxygen contamination and vice-versa. Water vapour in the ambient is responsible for the oxidation. The oxygen contamination significantly impacts the properties of the films. The E04 optical bandgap increases with oxygen contamination which hinders the development of a low-bandgap absorber layer, while the Eact decreases to values as low as around 50 eV . Generally, the pre-exponential factor (sigma_o) decreases significantly by 1-5 orders of magnitude which outweighs the decrease in Eact, resulting in the decrease in dark conductivity by 1-3 orders of magnitude. Consequently, highest photo/dark conductivity ratios of 5-6 are obtained for these films. Amongst the films without oxygen contamination, the lowest E04 optical bandgap reported is 1.2 eV, with a Etauc bandgap of 0.93 and a photo/dark conductivity ratio around 3.4. Most of the single-junction cells processed at absorber layer thicknesses of 100nm and above show resistor-like behaviour. The substantially high values of Rs losses associated to the Rsh degrade the cell parameters significantly. Although, for low absorber layer thickness of around 50nm, the closest resemblance to a cell-behaviour is observed. Therefore, there is a scope for improvement with regards to processing of these single-junction cells.

Single junction solar cells have a theoretical efficiency limit of 33.1% due to spectral mismatch. To overcome this limitation, two or more single junction cells with different bandgaps can be coupled together to achieve higher energy conversion efficiency by optimum spectral utilization. In this work, thin film silicon-based alloys are used to obtain a high potential and a current matched 2-junction solar cell is fabricated. The top cell is a-Si which has high bandgap of 1.8eV and the bottom cell is a-SiGe:H which has a tuneable bandgap between 1.4-1.6eV. This thesis will focus in optimizing the window layer of the top a-Si solar cell and the effects of Hydrogen Plasma Treatment (HPT) at the i/p and p/TCO interface. This top cell is fabricated in a PECVD cluster tool in n-i-p substrate configuration on ASAHI glass substrate. The use of PECVD allows for a better control of the layer properties by changing the gas flow rates and the deposition environment. Initially a reference cell is fabricated which had an open circuit voltage (Voc) of 804 mV, a fill factor (FF) of 0.63, a short circuit current density (Jsc) of 12.8 mAcm-2 and an efficiency of 6.56 %. To improve the performance of the solar cell the the effects of HPT on the material properties of the i-layer and the p-layer are studied at varying power, pressure and duration of exposure. The next step is to study the effects of the precursor gas flow rates for the deposition of p-nc-SiOx layer and optimize the thickness of the individual components of the window layer including the TCO. The optimization process involved trading off certain properties in favour of a high Voc and high FF. The final cell had an open circuit voltage (Voc) of 863V, a fill factor (FF) of 0.655, a short circuit current density (Jsc) of 12.5 mAcm-2 and an efficiency of 7.06 %. Thus, by only optimizing the non-active window layer an open circuit voltage gain of 60mV is achieved with an improved FF. This single junction cell is then fabricated on top of a-SiGe:H solar cell. With some iteration in the absorber layer properties of both the top and bottom cells, the final tandem cell had a Voc of 1395 mV, a FF of 0.69, the current limiting Jsc of the top cell at 8.34mAcm-2 and an overall efficiency of 7.99%. To further improve this tandem cell, the tunnel recombination junction can be further optimised.

This thesis is dedicated to the fabrication of a novel silicon based tandem cell which combines hydrogenated nanocrystalline silicon (nc-Si:H) thin film(TF) photovoltaic (PV) technology and silicon heterojunction (SHJ) c-Si based PV technology. As a matter of fact, the growth of nc-Si:H on a flat c-Si substrate is not uniformas it irregularly peels off after deposition using the plasma enhanced chemical vapor deposition (PECVD). On the other hand, the growth of the nc-Si:H TF layers on the standard alkaline textured c-Si substrate with sharp pyramidal structures results in defective regions in the bulk of nc-Si:H grown material. A nc-Si:H layer with high defect density reduces the voltage and the fill factor (FF) of the tandem device. In order to minimize the defect density in the nc-Si:H absorber layer, different texturing approaches (TA) were developed for the c-Si
wafer to facilitate better growth of the nc-Si:H absorber. The surface morphology of the textured c-Si wafers for all TA’s at different etching time steps were investigated. Prior to TF layer deposition for the SHJ and the tandem devices, various cleaning approaches were investigated to improve the surface passivation of the textured c-Si wafers. The investigation of the grown nc-Si:H layer using different TA’s showed that smoothening of the sharp pyramidal structures significantly improved the nc-Si:H grown bulk as it helped in better growth of nc-Si:H with no significant defects. Finally, the current density-voltage (J-V) measurements were investigated for all TA’s for the SHJ and the tandem devices. The best performing tandem cell has an open circuit voltage (Voc ) of 1.02 V , short current density (Jsc ) of 13.34 mA/cm-2 and a FF of 0.44. It is expected that optimizing the tunnel recombination junction (TRJ) will further improve the electrical performance of the tandem devices for all TA’s.