In this case study, the energy performance of a secondary school building from the 1960s in Gävle, Sweden, was modelled in the building energy simulation (BES) tool IDA ICE version 4.8 prior to major renovation planning. The objectives of the study were to validate the BES model during both occupied and unoccupied periods, investigate how to model airing and varying occupancy behaviour, and finally investigate energy use to identify potential energy-efficiency measures. The BES model was validated by using field measurements and evidence-based input. Thermal bridges, infiltration, mechanical ventilation, domestic hot water circulation losses, and space heating power were calculated and measured. A backcasting method was developed to model heat losses due to airing, opening windows and doors, and other occupancy behaviour through regression analysis between daily heat power and outdoor temperature. Validation results show good agreement: 3.4% discrepancy between space heating measurements and simulations during an unoccupied week. Corresponding monthly discrepancy varied between 5.5% and 10.6% during three months with occupants. Annual simulation indicates that the best potential renovation measures are changing to efficient windows, improved envelope airtightness, new controls of the HVAC system, and increased external wall thermal insulation.

The focus of this study is to assess the airflows in a school building built in 1963 in Gävle, Sweden, which is subject to energy conservation measures (ECMs) in a forthcoming renovation. Today, the school building is mainly ventilated by several mechanical ventilation systems, which are controlled by a constant air volume (CAV) strategy. Schedules and presence sensors impose a high operation mode during the day and a low operation mode at night, on weekends and on holidays. The homogeneous tracer gas emission method with passive sampling is used to measure the average local mean age of air (τ) during different operation modes. Temperature, relative humidity and CO₂ concentration are simultaneously measured. The calculated relative uncertainty for the average local mean age of air in every measured point is approx. ±20 %. The results during low operation mode show an average value of τ of approx. 8.51 h [corresponding to 0.12 air changes per hour (ACH)], where τ in various zones ranges between 2.55 and 16.37 h (indicating 0.06–0.39 ACH), which is related to the unintentional airflow in the school. The results during mixed operation mode show an average value of τ of approx. 4.60 h (0.22 ACH), where τ in various zones ranges between 2.00 and 8.98 h (0.11–0.50 ACH), which is related to both unintentional and intentional airflows in the school. Corridors, basement and attic rooms and entrances have lower τ compared to classrooms, offices and other rooms. High maximums of the CO₂ concentration in some rooms indicate an imbalance in the mechanical ventilation systems. During a regular school week of mixed operation, which includes both high and low operation modes, it is found that mainly the low operation modes show up in the results. The dynamics of the highly varying airflows in the building cannot be identified using the passive sampling technique.