Durham School of Architectural Engineering and Construction


Document Type


Date of this Version



Procedia Computer Science 86 ( 2016 ) 341 – 344, doi: 10.1016/j.procs.2016.05.092.


© 2016 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license.


Heating, ventilation, air-conditioning and Refrigeration (HVAC&R) systems are seldom designed or commissioned properly. The situation leads to abrupt or degradation faults resulting in inefficient energy uses, excessive energy consumption and high service costs. To solve these aforementioned problems, fault detection and diagnosis (FDD) is utilized to firstly detect any abnormal conditions of a system and then diagnoses and determines their causes. In order to apply this concept in HVAC oversizing designs, this paper proposes the state-of-art procedure of a FDD procedure for analyzing the inherently faulty design (oversizing) of multiple packaged air-conditioning units used to supply cooling for an open space in light commercial buildings. A generic process of FDD for a packaged unit is briefly introduced to efficiently design FDD algorithms and to illustrate an overview picture for new researchers in FDD areas. In the procedures, compressor statuses, time-on and time-off operations and outdoor air temperatures are recorded by means of the on-board controller of each machine unit. These physical and electrical monitoring data are applied to diagnose and evaluate oversizing level in terms of runtime fraction (RTF) and cycling rate (N). Eventually, an adaptive control is designed and implemented to enhance process recovery for soft-repairing and permanently reducing fault effect caused by oversizing without intervening system operations (non-invasive technology).