Solar PV systems have so far been the source of choice for the sustainable supply of urban electric transport networks—like trams and trolleybus grids. However, no consensus exists yet on the placement or sizing of PV systems at the traction substations, and no method is available for easy estimation of the PV system utilization performance. The latter is crucial for understanding the need for storage, grid exchange, or even power curtailment, and has therefore a direct impact on the technical and financial feasibility of the project. This paper looks at 11 Key Performance Indicators (KPI) that are available to trolleybus operators, in two PV case studies on Arnhem (NL) and Gdynia (PL), using verified and validated bus, grid, and PV models. Through one KPI, namely the here-defined Energy Traffic KPI, a strong trend (R²=0.93) is described that can now allow stakeholders a quick estimation of the PV potential using a simple third-degree polynomial instead of resorting to the complex grid, bus, and PV modelling. A simple placement and sizing method is also presented derived from this KPI, in a way as to increase the technical and economical feasibility of an installed PV system. Despite all efforts, stakeholders are still warned of an intrinsic, upper-performance plateau that exists in transport grids, at around 38% direct PV utilization, caused by the unavoidable mismatch between PV generation and vehicle timetables and schedules. Stakeholders are urged to implement more smart grid loads as a base load to increase the feasibility of their investments in renewables, and to transform the transportation systems thereby to multi-functional grids that can assist the main city grid.

This paper offers a complete and verified model of DC trolleybus grids and examines the effect of the common modelling assumptions made in literature by using simulations, as well as bus and substation measurements from the grid of Arnhem, the Netherlands. An equivalent model for the overhead line impedance is offered taking into account the single line impedance, the supply and return lines, and the parallel connections between them. A case study shows that the feeder cables from the substations to the sections can be ignored, but only for certain substation power and feeder-line length ranges. On the other hand, the often-neglected regenerative braking, bus auxiliaries load, bilateral connections, and the exact nominal substation voltage are found to be crucial for the correct modelling of a trolleybus grid.

Reducing the environmental impact of transportation requires the successful integration of renewable energy sources into the electrical transportation networks. However, the mismatch between renewable generation and the intermittent bus schedules causes temporary absence of loads and creates considerable excess energy, potentially rendering the systems economically infeasible. So far, studies on integration of renewables in transport grids were limited to decentralized solar PV systems (placed at the substation level), using statistical or simplified models, and concerned mainly with increasing the trolleygrid capacity. In this paper, both PV and Wind systems are considered and studied as to maximize their direct utilization by using verified simulation models for six different sizing and placement scenarios. The Dutch trolleygrid of Arnhem is used as a case study. Scenarios I to V looked at a decentralized renewable sources placement and ultimately concluded that PV systems at low-traffic substations are best sized for complete energy-neutrality, with daily storage systems. On the other hand, those at high-traffic substations should be without storage and sized below their energy-neutrality point — ideally, using the Marginal Utilization approach (scenario III). Finally, the Centralized (Aggregated) Energy-Neutral Wind and PV Approach of scenario VI offers the best outcome, with a hybrid solution of 53% PV and 47% Wind. This scenario offers a 54.1% direct bus load coverage. In comparison, scenario I, which had attempted a grid energy-neutrality in a decentralized manner, had only achieved 32.4% direct load coverage. The outcome of scenario VI can even be pushed to values above 80% by installing storage systems.