Handling Discrepancies in Building Reactive Management Using HAZOP and Diagnosis Analysis (ST-16-C022)

Complexities, with the modern smart buildings are continuously increasing with time. Building automation systems encompass many sensors and controllers to achieve the desired level of comfort. However, in reality it is difficult to achieve the perceived comfort because of different faults, occupant misusages and their consequences. Though, sometimes occupants do not well aware with the origin of different causes and their unusual impact on comfort and cost. In order to fill the gap between expectation or what was planned before and reality, building energy management systems (BEMS) need to be designed in such a way that they will be able to react or advise different reactive actions to the occupants. Considering real time scenarios all major anomalies primarily arises from three main sources: Different failures in buildings including HVAC, Misusage i.e. human irrelevant behaviors and Various unplanned situations related to occupancy and appliances. Above abnormal situations could cause huge penalty over occupant’s comfort and operational cost. Potential faults and their causes need to be identified. Present work considers a hazard and operability analysis (HAZOP) for the whole building system. HAZOP is used to perform a detail analysis of all possible causes of discrepancies in building operation. HAZOP is a qualitative approach but it can be quantified by using “Risk assessment matrix” based on the frequency and severity of faults. According to HAZOP analysis, building facility is divided in different sub-systems and each sub-system is studied in detail. Each sub-system assigned with a variable and deviations from the normal range of these variables are considered as symptoms. Considering the detected symptoms, a global signature table can be computed. However due to presence of non-isolable causes this table is further reduced to observable signature table. To avoid the decision error a minimal set of diagnosis is performed on the basis of difference between computed and observable signatures. Further, all the possible risks are ranked according to their degree of severity and integrated with reactive energy management. Finally, paper will address the application of the developed methodology for an office without HVAC.