Charging electric vehicles from solar energy provides a sustainable means of transportation. This paper shows the design of solar powered e-bike charging station that provides AC, DC and contactless charging of e-bikes. The DC charger allows direct DC charging of the e-bike from the DC power of the photovoltaic panels (PV) without the need for an external AC charger adapter. In case of the contactless charger, the bike can be charged without the use of any cables providing maximum convenience to the user. Finally, the charging station has an integrated battery that allows for both grid-connected and off-grid operation@en

Recent innovations in Electric Vehicles (EVs) will potentially change the future of the transportation industry. They will diversify the energy mix and reduce the dependence on fossil fuels. However, use of EVs only shifts the source of CO2 production to electricity generation plants. A smart solution to overcome this problem is the use of localized generated power and solar-powered charging stations are the best way to achieve it.

A solar powered e-bike charging station, installed on the TU Delft campus is one such example. The charging station is equipped with a meteorological station, sensors for monitoring performance, inverters and batteries. The PV system installed at the e-bike station was thoroughly modeled, considering both the location and meteorological conditions of the final installation [1]. To maximize the station’s utility, it is important to accurately predict the energy yield of the system. The modeling step comprises of several sub-models (irradiance, thermal and electrical model) which indicate the energy yield of the station as well as the power exchange with the grid. Though these models were based on (realistic) assumptions, there is a need to verify the assumptions against measured values.

In this thesis, the accuracy of existing irradiance, thermal and electrical models was evaluated by predicting the energy yield of the e-bike charging station. Further, the performance of these models, especially those related to the irradiance on the plane of the array and the instantaneous temperature of the PV modules, was improved. Also, two new decomposition models are introduced to improve the accuracy of obtaining diffuse irradiance from global horizontal irradiance specifically for the Netherlands. It was found that for accurate energy yield prediction it is necessary to optimize the models using
location specific parameters like sky view factor, albedo, INOCT etc. The energy yield predicted, using the improved models in this thesis, was only 17 kWℎ less than the measured yield for the duration Oct’16-Apr’17.