Simulation-based seismic loss assessment of RC school buildings using story-level damage ratio functions validated by observed cost data

School buildings are critical facilities during earthquakes; however, many existing reinforced concrete (RC) school buildings in Korea were constructed without seismic detailing, making them highly vulnerable to structural and nonstructural damage. Traditional drift-based performance assessment primarily focuses on life safety and structural damage criteria, while economic losses are typically not adopted as primary performance metrics for retrofit decision-making. This study aims to assess a loss-based seismic performance of RC school buildings that employs story-level damage ratio functions to estimate building-specific repair costs. The methodology integrates nonlinear time-history analysis, probabilistic seismic demand models using bilinear cloud analysis, and fragility functions for both structural and nonstructural components. Repair cost data are combined with fragility results to develop simulation-based seismic loss functions, which are correlated with drift limits to define loss-based performance criteria. Application to a representative RC school building demonstrated accurate simulation of observed damage patterns, including column shear failure and the influence of masonry infills. Regional-scale evaluations for the 2016 Gyeongju and 2017 Pohang earthquake scenarios confirmed the reliability of the approach, with estimated losses differing by 1.44 and 1.07 times from observed data, respectively. Results show that nonstructural components contribute over 70% of total losses, emphasizing their role in seismic risk. The proposed framework provides a practical tool for assessing economic vulnerability and supporting retrofit prioritization at both building and regional levels.