Are you an EPFL student looking for a semester project?
Work with us on data science and visualisation projects, and deploy your project as an app on top of Graph Search.
The current global energy landscape is characterized by an increasing demand for affordable and sustainable energy sources, leading to an ever increasing integration of intermittent renewable energy resources into the grid. The intermittent nature of these sources has led to a growing interest in energy storage technologies, which can help to smooth out fluctuations in supply and demand and improve the overall stability and resilience of the grid. Pumped hydro storage power plants are a proven technology for energy storage, with the ability to store large amounts of energy for long periods of time. However, their conventional fixed speed operation lacks the required flexibility to follow the fluctuations in energy production resulting from intermittent renewable energy sources. By utilizing the energy stored in pumped hydro storage facilities in a more flexible manner, we can reduce our reliance on fossil fuels and increase the use of renewable energy sources, helping to mitigate climate change and improve the sustainability of our energy systems. This flexibility can be achieved through variable speed operation.This thesis focuses on the use of the modular multilevel matrix converter in pumped hydro storage power plants to enable variable speed operation through the so-called converter fed synchronous machine configuration. The advantages of this variable speed configuration lie in its increased operating range and flexibility. The ratings of typical pumped hydro storage power plant units are within the ten to hundreds of MW combined with voltages in the tens of kV range. Achieving such ratings on the converter side requires emerging converter topologies such as the modular multilevel converter. Among this family of converters, the matrix topology outperforms the other topologies due to its better current sharing, which is further demonstrated in this thesis. Being a modular converter topology, due to its series connection of cells with individual energy storage elements, raises control challenges as to keep the energy within the converter balanced through all the cells. A proven method for the validation of the control structure and algorithm is the so-called real-time hardware-in-the-loop tests, which are widely used in industry. This thesis demonstrates how this complex converter topology can be divided into multiple sub-circuits, which can be deployed on various small scale real-time simulators to build a real-time hardware-in-the-loop platform. This platform is used throughout this thesis to generate high fidelity results regarding control validation and system optimization. This thesis provides an in-depth analysis of the three main control loops, namely the internal energy balancing, grid side current control and machine side current control. The currently existing energy control algorithms are brought into a common reference frame to allow a thorough mathematical comparison revealing major differences in the execution of the given task. The hardware-in-the-loop platform serves as tool to validate the mathematical findings, which allow multi-dimensional performance analysis. Being a grid connected storage facility, grid code requirements have to be followed, which is demonstrated under various scenarios in this thesis. The evaluation and implementation of the system losses allows for a system loss minimization by exploiting all degrees of freedom of the synchronous generator and constitutes another part of this thesis.
François Maréchal, Jonas Schnidrig
François Maréchal, Julia Granacher