Tandem technology has emerged as one of the most promising innovations in the field of photovoltaics. Higher conversion efficiencies than standard single-junction cells have already been achieved. Proving how these laboratory conversion capabilities translate into real-world performance is therefore a main interest drive within the photovoltaic research community.

Typically, photovoltaic device performance is assessed in Standard Test Conditions (STC). For the majority of climates, this is however not representative of actual operating conditions. With the aim of studying potential system applications, the energy yield prediction method adopted in this thesis analyzes tandem technology performance considering real-world conditions. For this scope, the PVMD Toolbox for PV system modeling is employed.

The tandem device investigated in this work is a perovskite/c-Si based structure in a two-terminal architecture. In particular, a TU Delft own poly-SiOx based crystalline silicon bottom cell is utilized in the configuration. This structure is implemented in the Toolbox to carry out Optical simulations from cell to module system level. Operating conditions are introduced through use of the Toolbox Weather and Thermal models. These recreate real-world illumination and climate settings for the selected locations of Reykjavik, Rome and Alice Springs. With an STC-optimized perovskite thickness of 515 nm, the hourly implied photocurrent density of the device is calculated and validated. The jph annual average is 1.92, 3.34 and 5.01 mA/cm2 for the three locations, respectively. These values take into account a sub-cells current mis-
match loss between 5 and 8% depending on location. To reproduce real-world cell electrical parameters behavior, the variable irradiance and temperature sub-cells J-V curves have been obtained. A poly-SiOx based silicon prototype was tested in laboratory settings and its curves then utilized in simulations after measurements validation.

The module annual energy yield is computed through simulations with the Toolbox Electric model. In terms of specific DC yield, the STC-optimized module delivers 820, 1,427 and 2,161 kWh/kWp/year in the three locations. The power mismatch lowering effect due to the fill factor gain compensates the current mismatch in this 2T configuration, reducing energy mismatch losses. This performance is then compared to a reference single-junction poly-SiOx based silicon module, to show how tandem technology potentially outperforms standard modules differently depending on climatic conditions. The tandem module is also assessed in terms of DC performance ratio (PR), exhibiting values over 0.94 for all three locations. The structure is then real-world optimized by varying perovskite thickness, with the goal of maximizing
location-specific energy yield and PR. Slight improvements are obtained for the Reykjavik and Alice Springs locations, where the energy yield increases by few relative percentage points along with PR.

Due to the high-efficiency potential over 40%, the two-terminal (2T) perovskite/c-Si tandem solar cell becomes a desirable and promising candidate for solar cells. The photovoltaic materials and devices (PVMD) group in TU Delft has developed the c-Si solar cell using poly-SiOx as carrier-selective passivating contacts. The utilization of poly-SiOx as carrier selective passivating contacts is relatively new and very interesting to be analyzed as its bandgap can be varied by changing its oxygen content. This can be used in the solar cell to increase the Jsc. In this work, the 2T perovskite/c-Si tandem solar cells' optimization has been conducted by means of optical and electrical simulations. Optical simulations of the 2T, 3T and 4T perovskite/c-Si tandem solar cells in GenPro4 are presented. The main objective is to show the new type of bottom cell's optical performance in the 2T, 3T, and 4T tandem solar cell. Furthermore, the two-terminal (2T) perovskite/c-Si tandem solar cell's electrical modeling framework with the new type of the bottom cell has been developed. In this model, the optical generation profile from GenPro4 was successfully imported into the semiconductor simulations software called Sentaurus TCAD. The perovskite and c-Si solar cells have been optimized separately. The perovskite solar cell optimization was conducted by changing parameters such as contact resistance, surface recombination velocity and thickness of each layer. This leads to the increase in the perovskite solar cell efficiency from 15.83% to 21.50%. This top cell was combined with a c-Si bottom cell with an efficiency of 25% to form a two-terminal (2T) tandem solar cell. The two-terminal (2T) perovskite/c-Si was further optimized by optimizing the tunnel recombination junction. The optimization of TRJ was done by varying the doping concentration of indium tin oxide (ITO) and spiro-OMeTAD, respectively. As a result, the 33.70% efficiency of the two-terminal (2T) perovskite/c-Si tandem solar cell was obtained.