As the cooling energy demand in buildings is set to increase dramatically in the future, the exploitation of passive solutions like natural ventilation could prove vital in reducing the reliance on mechanical systems. Models that can predict air temperature accurately in naturally ventilated mode are key to understanding the potential of natural ventilation now and in the future. This article presents a simulation based case study of a retrofitted nearly zero energy test-bed university building, in naturally ventilated mode only. The study had three aims: (1) calibration and validation of a whole building energy model, (2) a comparative analysis of occupancy schedules and opening control strategies, and (3) a comparison of researcher and practitioner approaches. Results showed the detailed building model was capable of predicting room level air temperature with a low level of error (0.27 degrees C <= 5 RMSE <= 1.50 degrees C) that was well within the limits of existing calibration standards (MBE +/- 10%, CVRMSE < 20%). The comparative analysis highlighted the need to consider occupancy schedules that have a wide range of diversity, and opening control strategies that reflect the manual and automated relationship in natural ventilation systems. The approach comparison highlighted that both practitioner and researcher approaches to simulating both occupancy schedules and opening control strategies showed similar levels of performance for the application considered. The paper also provides recommendations for those modelling air temperatures and thermal comfort in nearly zero energy buildings.