UC Merced

Heating, ventilation and air conditioning (HVAC) is a mechanical system that provides thermal comfort and acceptable indoor air quality. The HVAC system takes a dominant portion of overall building energy consumption and accounts for 50% of the energy used in the U.S. commercial and residential buildings in 2012. The performance and energy saving of building HVAC systems can be significantly improved by the implementation of better fault detection strategies.

Motivated by these goals, this thesis presents a scaled-up system level fault detection application based top and cross level fault detection schemes. Using top level and cross level schemes, energy consumption of devices at different levels and at the same level, is compared using principal component and correlation analysis respectively. Through these strategies, anomalies in energy consumption, which are indicators of faults are revealed. Moreover, energy consumption models are established for each type of device inside the system. These models are based on thermal and potential energy balances. This fault detection scheme forms the foundation of a fault detection program implemented in MATLAB that is easily adaptable to different types of HVAC systems.

Additionally, this thesis presents a methodology for organizing the data. The organizational structure of the data reflects the physical structure of the HVAC system. This structure facilitates data retrieval and application of spatial and temporal partitioning schemes.

In this thesis, all the data processing, models, and implementation of the fault detection program are based on extensive data measurements collected from an office building on the campus of the University of California, Merced.