Open window and defective radiator valve detection

The energy consumption of buildings is growing quickly. Because building auto-mation and HVAC systems can not run under optimal conditions and faults or disturbances can appear in the different phases, fault detection and diagnosis (FDD) is necessary. FDD can be used to trace the cause of the decrease of energy ef” ciency in an HVAC system or the decrease of the thermal comfort level in the rooms. This paper studies a model-based FDD in an of” ce room. By analysing the physical relations between all the components, the room models are built with SimulinkÒ. Two types of fault are studied in this paper: an open window and a defective radiator valve. By introducing a characteristic parameter, these two faults can be detected and diagnosed.