Abstract

The transition to net-zero transportation in urban areas requires an integrated understanding of clean electricity deployment, transit electrification strategies, and operational energy dynamics. This thesis presents an integrated analysis and modeling framework for evaluating energy demand, greenhouse gas (GHG) emissions, and life cycle impacts of electric bus deployment under low-carbon electricity scenarios in Canada. The research is structured into four interconnected studies, each addressing a critical layer of the energy transition.

The first study conducts a meta-analysis of 26 national decarbonization scenarios to evaluate the role of clean energy in achieving Canada's 2050 targets across electricity, transport, and heating sectors. Findings reveal consistent reliance on hydropower and nuclear (collectively ~80%), while the balance between wind, solar, and natural gas remains uncertain. The study highlights key challenges such as inter-provincial coordination and inadequate carbon pricing, particularly in the electricity sector.

Building on this, the second study proposes a regionalized modeling framework to assess electric bus penetration in Toronto, Montreal, Edmonton, and Halifax. Integrating capital costs, dynamic fuel prices, social costs of pollution, and carbon pricing, the study quantifies GHG reductions from 2019 to 2030, with emission cuts ranging from 18.7% to 34.6% under energy system decarbonization (ESD) scenarios. The results emphasize the need for localized strategies within a polycentric governance framework.

The third study analyzes real-world operational data from Montreal’s transit network to evaluate battery electric bus (BEB) energy performance under seasonal variations. Energy consumption peaks during winter due to auxiliary heating and road friction, while regenerative braking performs optimally at mid-speeds (30–50 km/h) in warmer conditions. Findings demonstrate the influence of climate and road conditions on BEB energy efficiency and operational cost.

The final study develops machine learning-based models to estimate trip-level BEB energy consumption using three years of operational data. The models incorporate external variables such as temperature, state of charge, traffic conditions, road gradient, and stop frequency. Key predictors of BEB energy demand are identified, offering insights for route optimization and energy-efficient scheduling. Together, these studies provide a multiscale, data-driven framework for understanding and advancing electric bus adoption in the context of clean energy transitions. The findings inform both national policy planning and local transit system optimization, supporting the broader goal of achieving sustainable, low-carbon urban mobility in cold-climate regions.