ASHRAE Guideline 36: High-performance sequences of operation (SOO) for Heating, Ventilation, and Air-conditioning (HVAC) Systems has been demonstrated to save 17%-30% energy under ideal simulation environments. However, HVAC systems are susceptible to various types of faults in a real building operation. There are no existing studies that pertain to a comprehensive fault impact analysis of the high-performance control sequences suggested by ASHRAE Guideline 36 for HVAC systems. How these sequences handle and adapt to the various types of faults is still largely unknown. In this context, a comprehensive fault impact analysis and robustness assessment of the high-performance control sequences is conducted. A Modelica-based medium office virtual testbed is developed following the air-side and the plant-side SOO. A total of 359 fault scenarios in three different seasonal operating conditions (cooling, shoulder, and heating seasons) are injected into the baseline model. The evaluated key performance indexes (KPIs) include the operational cost, source energy, site energy, control loop quality, thermal comfort, ventilation, and power system metrics. The faults of the most negative impact are identified for different seasonal operating conditions over all the KPIs. The results also show that high-performance control sequences are well adapted for the vast majority (∼90%) of all the fault scenarios over all the KPIs in this study.

ASHRAE 指南 36：供暖、通风和空调 (HVAC) 系统的高性能操作序列 (SOO) 已被证明在理想的模拟环境下可节省 17%-30% 的能源。然而，HVAC 系统在实际建筑操作中容易受到各种类型的故障的影响。没有现有的研究涉及对 HVAC 系统的 ASHRAE 指南 36 建议的高性能控制序列的全面故障影响分析。这些序列如何处理和适应各种类型的故障在很大程度上仍然未知。在这种情况下，对高性能控制序列进行了全面的故障影响分析和鲁棒性评估。遵循空气侧和工厂侧SOO开发了基于Modelica的中型办公室虚拟试验台。将三种不同季节工况（冷季、肩季和暖季）下的总共 359 个故障场景注入基线模型。评估的关键性能指标 (KPI) 包括运营成本、源能源、现场能源、控制回路质量、热舒适性、通风和电力系统指标。针对所有 KPI 的不同季节性操作条件，识别出最负面影响的故障。结果还表明，高性能控制序列适用于本研究中所有 KPI 的所有故障场景中的绝大多数（~90%）。控制回路质量、热舒适度、通风和电力系统指标。针对所有 KPI 的不同季节性操作条件，识别出最负面影响的故障。结果还表明，高性能控制序列适用于本研究中所有 KPI 的所有故障场景中的绝大多数（~90%）。控制回路质量、热舒适度、通风和电力系统指标。针对所有 KPI 的不同季节性操作条件，识别出最负面影响的故障。结果还表明，高性能控制序列适用于本研究中所有 KPI 的所有故障场景中的绝大多数（~90%）。