The need for proper fast charging infrastructures is one of the key challenges for the wide adoption of elec- trical vehicles (EV). The high pulsating demand of fast charging stations (FCS) together with high demand tariffs can cause monthly DSO demand charges to account for a significant fraction of a station’s electric bill. Therefore, weakening the business case for stations located in these high tariff regions. To tackle this issue, demand charge management (DCM) can be applied to suppress peak power demands at FCSs using battery energy storage systems (BESS). This enables the reduction of cost while retaining the station’s fast charging capabilities. However, the implementation of such systems remains a large investment and the proper BESS sizing in fast charging applications is not well studied. This thesis proposes a multi-objective approach for optimal BESS sizing at FCSs considering demand charges and station performance. A BESS assisted FCS model is formulated to analyse the performance of a station’s design based on power flow, charging delays and the expected BESS lifetime. Furthermore, based on a worst-case demand scenario, a multi-objective optimization framework is formulated using the genetic algorithm NSGA-II to obtain the optimal BESS and grid-tie sizing for an existing FCS. Lastly, with demand data measured at four FCSs in the Netherlands, a set of numerical case studies has been conducted in the Mosaik and Pymoo environments to assess the feasibility and the effectiveness of the proposed formulation. These case studies provide new insights on the demand charge reduction and optimal sizing regarding dif- ferent station characteristics, BESS prices, and demand tariffs. These insights show how the FCS utilization rate and installed capacity can effect the optimal BESS sizing and how different demand tariffs or BESS cost can result in different optimal power to energy (P/E) ratios, and thus affecting the performance of a BESS.

This document is the bachelor graduation thesis of BAP Group B2. Together with BAP Group
B1 the objective of this project was to create a control network and DC/DC converter implementation
capable of fast, analog power management with active, remote-control and -interfacing of
multiple converter’s energy flows. The system will be designed to function in the solar-powered
eBike charging station on the TU Delft campus.
This document specifically is concerned with the design, simulation and testing of the DC/DC
converter. The thesis presents a bi-directional flyback DC/DC converter design and prototype
with analog power management, universal control signals and high integration potential.
These aspects are accomplished by building upon the traditional flyback DC/DC converter design
with respect to circuit topology, power-flow control, connectivity and accessibility. The
difficult regulation and measurement of DC power flow has been reduced to a set of analog
signals interpretable by any microcontroller, easing the integration of the converter into larger
energy and control networks.
A working prototype is presented and the inner-workings of the electrical circuit are discussed
and evaluated in detail. The converter is tested beyond the limit of the design requirements and
the performance is documented. Provided are possible reasons for malfunctioning and guidelines
for future improvements.
Design