Modelling of detailed community interaction dynamics increases a public health organization's ability to contain a potential disease strain at an early stage. Due to its dense population and high levels of human movements and interactions, Hong Kong has suffered from various epidemic diseases. The use of non-medical interventions is often efficacious in containing pandemic outbreaks. In this paper, we focus on evaluating the effectiveness of various practical school-related non-medical intervention strategies to mitigate the effects of pandemic influenza under a realistic Hong Kong demographic scenario. We modelled the impact of a combination of various school closure modes, triggers, types, and lengths. The simulation results suggest that the strategy of closing all types of schools generally outperforms that of closing only a subset, especially if the closure period is substantial. We also discuss future research directions along with individual school closure and economic evaluations.