Condition Monitoring for Air Filters in HVAC systems with variable volume flow

State of the art condition monitoring systems for air filters in HVAC systems require that the HVAC system is operated at nominal volume flow. For HVAC systems with variable volume flow this assumption is only fulfilled in one operating point. Outside this operating point existing condition monitoring systems assess the air filter condition in a too optimistic manner. Therefore, polluted air filters remain undetected until their regular check, leading to unneeded energy consumption. If the true condition of an air filter is known, it could be changed before it is clogged. So, a condition monitoring systems is needed which is also reliable in case of HVAC systems with variable volume flow. This work presents a model-based approach for such a condition monitoring system. Therefore, a dataset from a building is used to assess an optimal model. Furthermore, the condition monitoring systems is evaluated on that dataset.