Model-based building performance evaluation and diagnosis

Effective and efficient performance diagnosis and evaluation at the building level is important. This thesis presents a systematic methodology for building performance evaluation and diagnosis viewing building system as a whole. This methodology involves the development of a building global performance evaluation tool and practical strategy and approach for practical applications using the evaluation tool.; The evaluation tool includes a simplified building energy model and a consolidation AHU model. The simplified building energy model consists of simplified models of building envelopes (3R2C) and a simplified model of internal mass (2R2C). The parameters of 3R2C models are identified using easily available physical properties of building envelopes based on frequency response characteristic analysis and genetic algorithm (GA) method. The parameters of 2R2C model are identified using short-term monitoring operation data based on GA method. The validation shows that the simplified model can dynamically and reliably predict the thermal energy consumption under different operation and weather conditions. The consolidation AHU model is developed to represent all the installed AHU systems for performance prediction of air side systems.; Based on the model prediction and site measurement, a performance signature-based diagnosis strategy is presented for performance diagnosis. Cooling energy consumption is chosen as an important indicator for energy performance diagnosis. The diagnosis process using the strategy shows that the causes resulting in the "measured" cooling energy consumption deviating can be identified qualitatively and quantitatively. The diagnosed causes are beneficial for further investigations of building system.; A performance evaluation approach is also presented to assess the energy efficiency of alternative control strategies for decision-making of building retrofitting and upgrading measures. The approach uses the building global performance evaluation tool for performance prediction. Then, an enthalpy bin method is used to divide predicted performance into a series of enthalpy bins to facilitate evaluation for practical applications. The test in a high rising commercial office building in Hong Kong shows that the approach can provide the best combination of alternative basis strategies for maximum electricity consumption saving.