Passive Crowd Buffering in Pre-Operational Transit Stations: A Hybrid Simulation Study of the Thammasat University SRT Red Line Station

Abstract

The planned Thammasat University station on the State Railway of Thailand's Red Line Phase II extension presents a pre-operational evaluation challenge common to transit-oriented stations serving university catchments: demand assumptions and resilience claims must be established before the station exists to validate them against. This study addresses that challenge through a dual-tier hybrid simulation, a macroscopic Machinations Discrete Event Simulation cross-validated against a microscopic AnyLogic pedestrian model, stress-tested across six scenarios spanning baseline operation, hardware degradation, network disruption, and impulse loading. The central finding is that the station's Skywalk Concourse functions as a passive crowd buffer: under both the most severe internal failure (escalator loss, 446 passengers) and an extreme exogenous load (a 1,500-passenger stadium egress, 415 passengers), the concourse absorbs spatial accumulation that would otherwise reach the platform, remaining within Pedestrian Level of Service B in every case. This buffering capacity, however, was tested against demand assumptions that this study finds to be conservative. Empirical validation against twelve months of origin-destination data from the existing thirteen-station network shows that all operating stations, including the two closest behavioural proxies for Thammasat, exceed the simulation's assumed peak-hour k-factors (12% AM, 10% PM) by 26 to 70%, with an empirically grounded range of 15 to 20% AM and 12 to 15% PM proposed for future scenario refinement. Station typology classification further shows that the two proxy stations fall into structurally different demand archetypes, indicating that network position rather than university-adjacency alone governs temporal demand concentration. Finally, Heat Index projections to 2050, anchored to an ERA5 baseline and a three-model CMIP6 ensemble, suggest a climatically-driven catchment-shrinkage mechanism that could render the concentrated-demand conditions already modelled closer to typical than exceptional by mid-century under high-emissions pathways. Together, these findings suggest that passive architectural width, not solely active crowd management, may represent a transferable resilience parameter for transit-oriented station design, warranting investigation across additional station typologies.