Design and modeling of a fuel cell hybrid eCargo bike prototype

Increasing concern about climate change, urban pollution, and traffic congestion has led some large cities, like Amsterdam and Utrecht, to ban completely combustion engines for inner city delivery vans within the next three years, a trend which will be likely followed by other metropolitan areas throughout Europe. To comply with this regulation trend, a few delivery companies have already started adopting eCargo bikes to perform last-mile delivery, harnessing the great potential of such vehicles in terms of operating costs and service quality. However, their diffusion in the parcel delivery sector is marginal because of their limited range. In fact, delivery eCargo bikes must travel 50 to 100 Km per day on average, which is very close to the 40 to 100 Km battery range declared by major eCargo bike manufacturers. This leads to poor real-world range performance, especially in hilly regions. In this context, the present work aims at designing a hydrogen fuel cell-based range extender to be installed on a commercial eCargo delivery bike and to evaluate its performances through simulation. A preliminary design was first carried out developing a low-fidelity, backward model of the original eCargo powertrain in order to find a commercial fuel cell meeting the average power demand over a reference drive cycle. Subsequently, appropriately sized hydrogen feed and purge circuit components were chosen, namely high pressure tank, pressure regulator, flowmeter, and all the necessary fittings to ensure a leak-free matching between the parts. Thereafter, the best layout of fuel cell, battery and power converters was determined such as to ensure the highest efficiency and an effective power split control between the two on-board energy sources. The resulting architecture included one buck DC-DC converter for the fuel cell auxiliaries, one buck-boost converter at the fuel cell output, and one bidirectional converter in charge of managing the battery and the power split. This phase was concluded with a proposed experimental layout to test the power electronics components in their assigned functions. Having defined vehicle architecture and components, a forward model of the eCargo bike prototype was finally created on Matlab/Simulink environment to study the bike performances under dynamic, more realistic conditions. Additionally, two fuzzy-logic based controllers were designed to manage power split and the regenerative braking torque. Results showed that a significant range extension of +145% in the most demanding conditions is achieved despite an additional 5 Kg of weight and about 10% less cargo volume. A slightly different behavior between the dynamic model and the backward simulation was observed and it was attributed to the power-split controller design, which has proven to be more successful in sustaining the battery charge level throughout the cycle rather than maximizing the fuel cell system efficiency.