Operable windows have become desirable design features of modern mechanically ventilated office buildings in North America. While they improve perceived control and adaptive comfort, their inappropriate use poses risks associated with increased heating and cooling energy use. Therefore, the sequence of operations for terminal devices serving zones with operable windows should be designed in recognition of these risks, which in turn should be informed by research investigating occupants’ window and thermostat use behavior. To this end, this paper examines window and thermostat use data collected from two mixed-mode ventilation buildings in Ottawa, Canada. Discrete-time Markov logistic regression models and decision tree models were established to predict the likelihood of thermostat keypress and window opening/closing instances and identify the indoor conditions that trigger these actions. Based on this analysis, a set of preliminary recommendations is developed to improve terminal device sequencing in mixed-mode ventilation buildings in cold climates such that the comfort and energy savings potential of operable windows can be fully realized. The recommendations include applying thermostat setpoint setback to encourage occupants to open windows when conditions are advantageous for saving energy and discourage occupants from opening windows when energy penalties may be caused.

可操作的窗户已成为北美现代机械通风办公楼的理想设计特色。虽然它们改善了感知控制和适应性舒适度，但它们的不当使用会带来与增加加热和冷却能源使用相关的风险。因此，应根据这些风险来设计服务于具有可操作窗户的区域的终端设备的操作顺序，而这些风险又应通过对居住者窗户和恒温器使用行为的研究来了解。为此，本文研究了从加拿大渥太华的两座混合模式通风建筑收集的窗户和恒温器使用数据。建立离散时间马尔可夫逻辑回归模型和决策树模型来预测恒温器按键和窗户打开/关闭实例的可能性，并识别触发这些动作的室内条件。在此分析的基础上，提出了一套初步建议，以改善寒冷气候下混合模式通风建筑的终端设备排序，从而充分发挥可操作窗户的舒适性和节能潜力。这些建议包括应用恒温器设定点回退，以鼓励居住者在有利于节能的条件下打开窗户，并在可能导致能源损失时阻止居住者打开窗户。制定了一套初步建议，以改善寒冷气候下混合模式通风建筑的终端设备顺序，从而充分发挥可操作窗户的舒适性和节能潜力。这些建议包括应用恒温器设定点回退，以鼓励居住者在有利于节能的条件下打开窗户，并在可能导致能源损失时阻止居住者打开窗户。制定了一套初步建议，以改善寒冷气候下混合模式通风建筑的终端设备顺序，从而充分发挥可操作窗户的舒适性和节能潜力。这些建议包括应用恒温器设定点回退，以鼓励居住者在有利于节能的条件下打开窗户，并在可能导致能源损失时阻止居住者打开窗户。