• During the COVID-19 pandemic, provincial governments in Canada were faced with the problem of allocating a limited vaccine supply. Governments had to decide how to prioritize individuals, and they based their decisions on susceptibility, transmissibility, severity, and social disadvantage. Our work focuses on modelling the population of Newfoundland and Labrador (NL) using agent-based simulation (ABS) to produce a social contact network that helps inform vaccine allocation decisions. We use mixed-integer programming to generate a synthetic agent population from Census data, simulate social interaction using an ABS model previously studied for pandemic intervention effectiveness, and solve a graph optimization problem that maximally disconnects social contact networks to mitigate disease spread. Further, we aim to investigate vaccination strategies that prioritize socially disadvantaged populations to address disparate health outcomes caused by pandemics.