SuBus Optimizes Depot Charger Sizing and Bus Charging Schedules with a Pattern-Based Mixed-Integer Model

Faced with growing pressure to reduce polluting emissions in public transport, this paper presents the experience of SuBus, a bus operator in Santiago de Chile, which decided to replace its diesel fleet with electric buses. This company operates around 1,000 buses and controls about 15% of the city of Santiago’s commercial kilometers. We aimed to determine the optimal number of electric chargers and schedule their daily use to minimize operating costs while preserving service quality. To address this challenge, we build a pattern-based mixed-integer linear programming model enumerating charging patterns, incorporating constraints on bus autonomy, terminal capacity, charging windows, and maintenance requirements. We feed the optimization model with real data from the company, including route scheduling, fleet and loader technical specifications, and effective public transport demand. Applying this approach enabled significant reductions in the number of chargers required and the charging frequency per vehicle. This resulted in lower investment and operating costs, more efficient use of existing infrastructure, and longer battery life. Compared with the previously used manual approach, our solution enables more efficient and sustainable decision making. This case shows how advanced analytical techniques transform traditional operational planning, enhance competitiveness, and enable rapid response to technological challenges in the sector.

History: This paper was refereed.

Funding: C.E. Cortés acknowledges partial funding from project ANID/FONDECYT/REGULAR 1241888.