Promotie / Defence
Auditorium De Tweede Hoofdwet
Kasteelpark Arenberg 41 3001 Heverlee
Following the corona regulations, the maximum capacity of the auditorium is 15 attendees +
jury. You are more than welcome to participate in person or online. In both cases, please register here.
Promotor / Supervisor
Prof. dr. ir. J. Driesen
Samenvatting van het onderzoek / Summary of Research
PhotoVoltaic (PV) generation has a lot of momentum nowadays. The use of PV modules for electricity generation is driven by economic reasons such as decreasing module prices, the easyness to scale and low maintenance. Also the increased public awareness regarding climate change accelerates the strive of governments and individuals for more renewable energy sources, such as PV, in the energy mix.
Due to the easyness to scale PV installations, it is found in both very large utility-scale systems (PV farms) as well as in small-scale installations around buildings. The most common form of PV generation in the built environment is Building-Applied PV (BAPV), where PV modules are placed on an existing building. A common BAPV example are modules that are placed on an existing roof. However, alternatives exist where the PV cells are an integral part of the construction elements. This technology is referred to as Building-Integrated PV (BIPV). Tesla's Solar Roof tiles are a well-known example of BIPV although a variety of forms exists such as shaders, parapets and façade modules, all covered by the term BIPV. The use of BIPV is driven by cost savings, as multiple functions can be combined in one product; by a greener image, e.g. when a company wants to highlight their efforts for climate change; and by policy, e.g. the European Performance on Buildings Directive (EPBD) which requires that all new buildings from 2020 on are Near Zero Energy Buildings (NZEBs). This requirement can be challenging for tall buildings, where the available roof space for PV is scarce compared to the building's energy demand. The use of façade BIPV then comes in handy to meet the
Up to now, the electrical installation of BIPV systems is typically done using a string inverter, similar to BAPV systems. In this thesis, that approach is questioned. More specifically, this thesis investigates how the electrical architecture of BIPV modules should be implemented to achieve a high energetic efficiency, a high level of safety, a long lifetime, at the lowest possible cost.
In the first chapter, the rationale behind the thesis is described in more detail and an overview of the research questions of each chapter is also included.
In the second chapter, different electrical architectures such as string inverters, micro-inverters, series power optimizers and parallel power optimizers are compared and held against key performance indicators of BIPV installations. The parallel power optimizer, hereafter referred to as a Module Level Converter (MLC), approach turns out to be the best solution, especially for BIPV façades where partial shading occurs. However, a market research revealed that the commercially available MLCs do not comply with BIPV requirements in terms of dimensions and electro-thermal properties. The third chapter focuses on safety aspects of the electrical installation. The MLCs are connected together on a Low-
Voltage DC (LVDC) bus. Converter controlled LVDC systems are a relatively new phenomenon and the fault behaviour of these systems requires further investigation. In this thesis, contact voltages for different grounding strategies are derived and a selective protection methodology is proposed and experimentally validated.
Two important parameters that influence the system efficiency are the DC voltage level and the type of solar cells. The influence of these parameters is investigated for two case study buildings in chapter four by the development of an electrical loss model that comprises the losses in the MLC, in the cabling and in the voltage balancing converter. The model is ran using one-day experimental data and shows that not only the losses but also the loss distribution is strongly affected by the DC voltage and the PV technology.
In the fifth chapter, an alternative for regular cabling in BIPV systems is proposed. Using laminated bus bars benefits the voltage stability of the system and allows to reduce the MLC output capacitance, leading to reduced costs and increased reliability. An experimental prototype was designed and tested to highlight the electrical advantages and lead to the insight that the interconnection between the modules is an important design aspect to use the full potential of a bus bar system.
The sixth and seventh chapter discuss the influence of respectively components and control on converter reliability. An overview of recommendations is presented for the selection of components and topologies. Furthermore, MLC prototypes that were developed during this thesis are discussed in detail. Finally, the impact of control on transistor lifetime is investigated using a one-year experimental dataset to highlight the influence of the conduction mode and the phase-shedding control strategy.
Examencommissie / Board of examiners
Em. prof. dr. ir. Jean-Pierre Celis, chairman
Prof. dr. ir. Arch. Dirk Saelens
Prof. dr. ir. Jozef Poortmans
Prof. dr. ir. Jan Cappelle
Prof. dr. ir. Michaël Daenen
Prof. dr. ir. Patrizio Manganiello